{
"cells": [
{
"cell_type": "markdown",
"id": "89da7014-4a27-423f-99cc-a59c716fe72d",
"metadata": {},
"source": [
"# Diabetes Prediction - Classification"
]
},
{
"cell_type": "markdown",
"id": "46d63423-ed48-42c4-ba9c-4f9ef23e3163",
"metadata": {},
"source": [
" "
]
},
{
"cell_type": "markdown",
"id": "722a587f-9099-4e21-95a9-433d07c3e458",
"metadata": {},
"source": [
"### Dataset Download \n",
"You can download the CSV file here: \n",
"[https://www.kaggle.com/competitions/playground-series-s5e12/data)"
]
},
{
"cell_type": "markdown",
"id": "7f621557-9928-475a-8bcf-2f7442abcc3d",
"metadata": {},
"source": [
"### Introduction\n",
"This project aims to predict the probability of a patient being diagnosed with diabetes using demographic, lifestyle, and health-related features.\n",
"Exploratory data analysis and rule-based feature engineering were applied.."
]
},
{
"cell_type": "markdown",
"id": "5b8fb302-ea95-44ea-b188-4fec58ada4f7",
"metadata": {},
"source": [
"### Import Libraries"
]
},
{
"cell_type": "code",
"execution_count": 80,
"id": "1479890b-e10e-4c09-99c8-ce41f8c30136",
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"import seaborn as sns\n",
"from sklearn.model_selection import train_test_split\n",
"from sklearn.metrics import accuracy_score, confusion_matrix, classification_report\n",
"from sklearn.linear_model import LogisticRegression\n",
"from lightgbm import LGBMClassifier\n",
"from xgboost import XGBClassifier\n",
"from sklearn.metrics import roc_auc_score\n",
"from sklearn.metrics import roc_curve,auc\n",
"import joblib\n",
"import warnings\n",
"warnings.filterwarnings('ignore')\n",
"pd.set_option('display.max_columns',100)"
]
},
{
"cell_type": "markdown",
"id": "019f9ae9-ed68-45de-a1b0-09f66969d1d1",
"metadata": {},
"source": [
"### Load Data"
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "6d2d9889-dce4-42c2-92ac-be46538b8d58",
"metadata": {},
"outputs": [],
"source": [
"df1=pd.read_csv('train.csv')"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "f1158f6d-f7fa-433a-b395-4532c210abeb",
"metadata": {},
"outputs": [],
"source": [
"df2=pd.read_csv('test.csv')"
]
},
{
"cell_type": "markdown",
"id": "ca7b7fed-b25d-427e-a957-08966bab3fe4",
"metadata": {},
"source": [
"### EDA"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "902d474b-1379-4c6c-8345-1a1a781f8db7",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" age \n",
" alcohol_consumption_per_week \n",
" physical_activity_minutes_per_week \n",
" diet_score \n",
" sleep_hours_per_day \n",
" screen_time_hours_per_day \n",
" bmi \n",
" waist_to_hip_ratio \n",
" systolic_bp \n",
" diastolic_bp \n",
" heart_rate \n",
" cholesterol_total \n",
" hdl_cholesterol \n",
" ldl_cholesterol \n",
" triglycerides \n",
" gender \n",
" ethnicity \n",
" education_level \n",
" income_level \n",
" smoking_status \n",
" employment_status \n",
" family_history_diabetes \n",
" hypertension_history \n",
" cardiovascular_history \n",
" diagnosed_diabetes \n",
" \n",
" \n",
" \n",
" \n",
" 0 \n",
" 0 \n",
" 31 \n",
" 1 \n",
" 45 \n",
" 7.7 \n",
" 6.8 \n",
" 6.1 \n",
" 33.4 \n",
" 0.93 \n",
" 112 \n",
" 70 \n",
" 62 \n",
" 199 \n",
" 58 \n",
" 114 \n",
" 102 \n",
" Female \n",
" Hispanic \n",
" Highschool \n",
" Lower-Middle \n",
" Current \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" \n",
" \n",
" 1 \n",
" 1 \n",
" 50 \n",
" 2 \n",
" 73 \n",
" 5.7 \n",
" 6.5 \n",
" 5.8 \n",
" 23.8 \n",
" 0.83 \n",
" 120 \n",
" 77 \n",
" 71 \n",
" 199 \n",
" 50 \n",
" 121 \n",
" 124 \n",
" Female \n",
" White \n",
" Highschool \n",
" Upper-Middle \n",
" Never \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" \n",
" \n",
" 2 \n",
" 2 \n",
" 32 \n",
" 3 \n",
" 158 \n",
" 8.5 \n",
" 7.4 \n",
" 9.1 \n",
" 24.1 \n",
" 0.83 \n",
" 95 \n",
" 89 \n",
" 73 \n",
" 188 \n",
" 59 \n",
" 114 \n",
" 108 \n",
" Male \n",
" Hispanic \n",
" Highschool \n",
" Lower-Middle \n",
" Never \n",
" Retired \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 0.0 \n",
" \n",
" \n",
" 3 \n",
" 3 \n",
" 54 \n",
" 3 \n",
" 77 \n",
" 4.6 \n",
" 7.0 \n",
" 9.2 \n",
" 26.6 \n",
" 0.83 \n",
" 121 \n",
" 69 \n",
" 74 \n",
" 182 \n",
" 54 \n",
" 85 \n",
" 123 \n",
" Female \n",
" White \n",
" Highschool \n",
" Lower-Middle \n",
" Current \n",
" Employed \n",
" 0 \n",
" 1 \n",
" 0 \n",
" 1.0 \n",
" \n",
" \n",
" 4 \n",
" 4 \n",
" 54 \n",
" 1 \n",
" 55 \n",
" 5.7 \n",
" 6.2 \n",
" 5.1 \n",
" 28.8 \n",
" 0.90 \n",
" 108 \n",
" 60 \n",
" 85 \n",
" 206 \n",
" 49 \n",
" 131 \n",
" 124 \n",
" Male \n",
" White \n",
" Highschool \n",
" Upper-Middle \n",
" Never \n",
" Retired \n",
" 0 \n",
" 1 \n",
" 0 \n",
" 1.0 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id age alcohol_consumption_per_week physical_activity_minutes_per_week \\\n",
"0 0 31 1 45 \n",
"1 1 50 2 73 \n",
"2 2 32 3 158 \n",
"3 3 54 3 77 \n",
"4 4 54 1 55 \n",
"\n",
" diet_score sleep_hours_per_day screen_time_hours_per_day bmi \\\n",
"0 7.7 6.8 6.1 33.4 \n",
"1 5.7 6.5 5.8 23.8 \n",
"2 8.5 7.4 9.1 24.1 \n",
"3 4.6 7.0 9.2 26.6 \n",
"4 5.7 6.2 5.1 28.8 \n",
"\n",
" waist_to_hip_ratio systolic_bp diastolic_bp heart_rate \\\n",
"0 0.93 112 70 62 \n",
"1 0.83 120 77 71 \n",
"2 0.83 95 89 73 \n",
"3 0.83 121 69 74 \n",
"4 0.90 108 60 85 \n",
"\n",
" cholesterol_total hdl_cholesterol ldl_cholesterol triglycerides gender \\\n",
"0 199 58 114 102 Female \n",
"1 199 50 121 124 Female \n",
"2 188 59 114 108 Male \n",
"3 182 54 85 123 Female \n",
"4 206 49 131 124 Male \n",
"\n",
" ethnicity education_level income_level smoking_status employment_status \\\n",
"0 Hispanic Highschool Lower-Middle Current Employed \n",
"1 White Highschool Upper-Middle Never Employed \n",
"2 Hispanic Highschool Lower-Middle Never Retired \n",
"3 White Highschool Lower-Middle Current Employed \n",
"4 White Highschool Upper-Middle Never Retired \n",
"\n",
" family_history_diabetes hypertension_history cardiovascular_history \\\n",
"0 0 0 0 \n",
"1 0 0 0 \n",
"2 0 0 0 \n",
"3 0 1 0 \n",
"4 0 1 0 \n",
"\n",
" diagnosed_diabetes \n",
"0 1.0 \n",
"1 1.0 \n",
"2 0.0 \n",
"3 1.0 \n",
"4 1.0 "
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df1.head()"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "b03545d7-08f3-42ab-971f-ea53b162e739",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" age \n",
" alcohol_consumption_per_week \n",
" physical_activity_minutes_per_week \n",
" diet_score \n",
" sleep_hours_per_day \n",
" screen_time_hours_per_day \n",
" bmi \n",
" waist_to_hip_ratio \n",
" systolic_bp \n",
" diastolic_bp \n",
" heart_rate \n",
" cholesterol_total \n",
" hdl_cholesterol \n",
" ldl_cholesterol \n",
" triglycerides \n",
" gender \n",
" ethnicity \n",
" education_level \n",
" income_level \n",
" smoking_status \n",
" employment_status \n",
" family_history_diabetes \n",
" hypertension_history \n",
" cardiovascular_history \n",
" diagnosed_diabetes \n",
" \n",
" \n",
" \n",
" \n",
" 699995 \n",
" 699995 \n",
" 29 \n",
" 1 \n",
" 59 \n",
" 6.9 \n",
" 5.2 \n",
" 1.5 \n",
" 26.1 \n",
" 0.88 \n",
" 133 \n",
" 57 \n",
" 69 \n",
" 163 \n",
" 58 \n",
" 90 \n",
" 126 \n",
" Female \n",
" Hispanic \n",
" Postgraduate \n",
" Upper-Middle \n",
" Former \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 0.0 \n",
" \n",
" \n",
" 699996 \n",
" 699996 \n",
" 46 \n",
" 2 \n",
" 72 \n",
" 7.7 \n",
" 7.7 \n",
" 3.8 \n",
" 25.5 \n",
" 0.85 \n",
" 106 \n",
" 85 \n",
" 65 \n",
" 188 \n",
" 45 \n",
" 107 \n",
" 119 \n",
" Female \n",
" Hispanic \n",
" Graduate \n",
" Upper-Middle \n",
" Former \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 1 \n",
" 1.0 \n",
" \n",
" \n",
" 699997 \n",
" 699997 \n",
" 35 \n",
" 1 \n",
" 50 \n",
" 5.6 \n",
" 6.1 \n",
" 6.4 \n",
" 26.9 \n",
" 0.88 \n",
" 127 \n",
" 84 \n",
" 63 \n",
" 168 \n",
" 59 \n",
" 77 \n",
" 166 \n",
" Female \n",
" White \n",
" Graduate \n",
" Middle \n",
" Never \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" \n",
" \n",
" 699998 \n",
" 699998 \n",
" 49 \n",
" 2 \n",
" 70 \n",
" 5.7 \n",
" 6.9 \n",
" 4.7 \n",
" 25.2 \n",
" 0.86 \n",
" 116 \n",
" 67 \n",
" 69 \n",
" 198 \n",
" 55 \n",
" 108 \n",
" 133 \n",
" Female \n",
" White \n",
" Highschool \n",
" Lower-Middle \n",
" Never \n",
" Retired \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" \n",
" \n",
" 699999 \n",
" 699999 \n",
" 42 \n",
" 4 \n",
" 56 \n",
" 4.2 \n",
" 7.0 \n",
" 6.1 \n",
" 28.3 \n",
" 0.90 \n",
" 99 \n",
" 73 \n",
" 74 \n",
" 191 \n",
" 63 \n",
" 105 \n",
" 159 \n",
" Female \n",
" Black \n",
" Graduate \n",
" Low \n",
" Former \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id age alcohol_consumption_per_week \\\n",
"699995 699995 29 1 \n",
"699996 699996 46 2 \n",
"699997 699997 35 1 \n",
"699998 699998 49 2 \n",
"699999 699999 42 4 \n",
"\n",
" physical_activity_minutes_per_week diet_score sleep_hours_per_day \\\n",
"699995 59 6.9 5.2 \n",
"699996 72 7.7 7.7 \n",
"699997 50 5.6 6.1 \n",
"699998 70 5.7 6.9 \n",
"699999 56 4.2 7.0 \n",
"\n",
" screen_time_hours_per_day bmi waist_to_hip_ratio systolic_bp \\\n",
"699995 1.5 26.1 0.88 133 \n",
"699996 3.8 25.5 0.85 106 \n",
"699997 6.4 26.9 0.88 127 \n",
"699998 4.7 25.2 0.86 116 \n",
"699999 6.1 28.3 0.90 99 \n",
"\n",
" diastolic_bp heart_rate cholesterol_total hdl_cholesterol \\\n",
"699995 57 69 163 58 \n",
"699996 85 65 188 45 \n",
"699997 84 63 168 59 \n",
"699998 67 69 198 55 \n",
"699999 73 74 191 63 \n",
"\n",
" ldl_cholesterol triglycerides gender ethnicity education_level \\\n",
"699995 90 126 Female Hispanic Postgraduate \n",
"699996 107 119 Female Hispanic Graduate \n",
"699997 77 166 Female White Graduate \n",
"699998 108 133 Female White Highschool \n",
"699999 105 159 Female Black Graduate \n",
"\n",
" income_level smoking_status employment_status \\\n",
"699995 Upper-Middle Former Employed \n",
"699996 Upper-Middle Former Employed \n",
"699997 Middle Never Employed \n",
"699998 Lower-Middle Never Retired \n",
"699999 Low Former Employed \n",
"\n",
" family_history_diabetes hypertension_history cardiovascular_history \\\n",
"699995 0 0 0 \n",
"699996 0 0 1 \n",
"699997 0 0 0 \n",
"699998 0 0 0 \n",
"699999 0 0 0 \n",
"\n",
" diagnosed_diabetes \n",
"699995 0.0 \n",
"699996 1.0 \n",
"699997 1.0 \n",
"699998 1.0 \n",
"699999 1.0 "
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df1.tail()"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "6fae2006-8d6d-42cc-b8a5-ae1050a02a1a",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" age \n",
" alcohol_consumption_per_week \n",
" physical_activity_minutes_per_week \n",
" diet_score \n",
" sleep_hours_per_day \n",
" screen_time_hours_per_day \n",
" bmi \n",
" waist_to_hip_ratio \n",
" systolic_bp \n",
" diastolic_bp \n",
" heart_rate \n",
" cholesterol_total \n",
" hdl_cholesterol \n",
" ldl_cholesterol \n",
" triglycerides \n",
" gender \n",
" ethnicity \n",
" education_level \n",
" income_level \n",
" smoking_status \n",
" employment_status \n",
" family_history_diabetes \n",
" hypertension_history \n",
" cardiovascular_history \n",
" \n",
" \n",
" \n",
" \n",
" 0 \n",
" 700000 \n",
" 45 \n",
" 4 \n",
" 100 \n",
" 4.3 \n",
" 6.8 \n",
" 6.2 \n",
" 25.5 \n",
" 0.84 \n",
" 123 \n",
" 70 \n",
" 64 \n",
" 209 \n",
" 55 \n",
" 135 \n",
" 111 \n",
" Female \n",
" White \n",
" Highschool \n",
" Middle \n",
" Former \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
" 1 \n",
" 700001 \n",
" 35 \n",
" 1 \n",
" 87 \n",
" 3.5 \n",
" 4.6 \n",
" 9.0 \n",
" 28.6 \n",
" 0.88 \n",
" 120 \n",
" 74 \n",
" 59 \n",
" 159 \n",
" 47 \n",
" 83 \n",
" 145 \n",
" Female \n",
" White \n",
" Highschool \n",
" Middle \n",
" Never \n",
" Unemployed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
" 2 \n",
" 700002 \n",
" 45 \n",
" 1 \n",
" 61 \n",
" 7.6 \n",
" 6.8 \n",
" 7.0 \n",
" 28.5 \n",
" 0.94 \n",
" 112 \n",
" 71 \n",
" 75 \n",
" 173 \n",
" 43 \n",
" 99 \n",
" 184 \n",
" Male \n",
" White \n",
" Highschool \n",
" Low \n",
" Never \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
" 3 \n",
" 700003 \n",
" 55 \n",
" 2 \n",
" 81 \n",
" 7.3 \n",
" 7.3 \n",
" 5.0 \n",
" 26.9 \n",
" 0.91 \n",
" 114 \n",
" 81 \n",
" 61 \n",
" 203 \n",
" 59 \n",
" 116 \n",
" 128 \n",
" Male \n",
" White \n",
" Graduate \n",
" Middle \n",
" Former \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
" 4 \n",
" 700004 \n",
" 77 \n",
" 2 \n",
" 29 \n",
" 7.3 \n",
" 7.6 \n",
" 8.5 \n",
" 22.0 \n",
" 0.83 \n",
" 131 \n",
" 78 \n",
" 79 \n",
" 177 \n",
" 59 \n",
" 87 \n",
" 133 \n",
" Male \n",
" White \n",
" Graduate \n",
" Low \n",
" Current \n",
" Unemployed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id age alcohol_consumption_per_week \\\n",
"0 700000 45 4 \n",
"1 700001 35 1 \n",
"2 700002 45 1 \n",
"3 700003 55 2 \n",
"4 700004 77 2 \n",
"\n",
" physical_activity_minutes_per_week diet_score sleep_hours_per_day \\\n",
"0 100 4.3 6.8 \n",
"1 87 3.5 4.6 \n",
"2 61 7.6 6.8 \n",
"3 81 7.3 7.3 \n",
"4 29 7.3 7.6 \n",
"\n",
" screen_time_hours_per_day bmi waist_to_hip_ratio systolic_bp \\\n",
"0 6.2 25.5 0.84 123 \n",
"1 9.0 28.6 0.88 120 \n",
"2 7.0 28.5 0.94 112 \n",
"3 5.0 26.9 0.91 114 \n",
"4 8.5 22.0 0.83 131 \n",
"\n",
" diastolic_bp heart_rate cholesterol_total hdl_cholesterol \\\n",
"0 70 64 209 55 \n",
"1 74 59 159 47 \n",
"2 71 75 173 43 \n",
"3 81 61 203 59 \n",
"4 78 79 177 59 \n",
"\n",
" ldl_cholesterol triglycerides gender ethnicity education_level \\\n",
"0 135 111 Female White Highschool \n",
"1 83 145 Female White Highschool \n",
"2 99 184 Male White Highschool \n",
"3 116 128 Male White Graduate \n",
"4 87 133 Male White Graduate \n",
"\n",
" income_level smoking_status employment_status family_history_diabetes \\\n",
"0 Middle Former Employed 0 \n",
"1 Middle Never Unemployed 0 \n",
"2 Low Never Employed 0 \n",
"3 Middle Former Employed 0 \n",
"4 Low Current Unemployed 0 \n",
"\n",
" hypertension_history cardiovascular_history \n",
"0 0 0 \n",
"1 0 0 \n",
"2 0 0 \n",
"3 0 0 \n",
"4 0 0 "
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df2.head()"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "09b4b197-3465-4991-b4fa-eb50b375ecef",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" age \n",
" alcohol_consumption_per_week \n",
" physical_activity_minutes_per_week \n",
" diet_score \n",
" sleep_hours_per_day \n",
" screen_time_hours_per_day \n",
" bmi \n",
" waist_to_hip_ratio \n",
" systolic_bp \n",
" diastolic_bp \n",
" heart_rate \n",
" cholesterol_total \n",
" hdl_cholesterol \n",
" ldl_cholesterol \n",
" triglycerides \n",
" gender \n",
" ethnicity \n",
" education_level \n",
" income_level \n",
" smoking_status \n",
" employment_status \n",
" family_history_diabetes \n",
" hypertension_history \n",
" cardiovascular_history \n",
" \n",
" \n",
" \n",
" \n",
" 299995 \n",
" 999995 \n",
" 59 \n",
" 3 \n",
" 185 \n",
" 6.3 \n",
" 7.3 \n",
" 4.4 \n",
" 22.8 \n",
" 0.81 \n",
" 108 \n",
" 75 \n",
" 64 \n",
" 189 \n",
" 54 \n",
" 96 \n",
" 126 \n",
" Male \n",
" White \n",
" Highschool \n",
" Upper-Middle \n",
" Former \n",
" Employed \n",
" 1 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
" 299996 \n",
" 999996 \n",
" 50 \n",
" 2 \n",
" 25 \n",
" 5.8 \n",
" 7.8 \n",
" 4.5 \n",
" 29.6 \n",
" 0.93 \n",
" 112 \n",
" 70 \n",
" 78 \n",
" 158 \n",
" 43 \n",
" 84 \n",
" 112 \n",
" Male \n",
" Asian \n",
" Postgraduate \n",
" Lower-Middle \n",
" Never \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
" 299997 \n",
" 999997 \n",
" 63 \n",
" 1 \n",
" 252 \n",
" 5.2 \n",
" 7.5 \n",
" 8.5 \n",
" 25.1 \n",
" 0.77 \n",
" 129 \n",
" 63 \n",
" 57 \n",
" 168 \n",
" 33 \n",
" 105 \n",
" 135 \n",
" Female \n",
" White \n",
" Highschool \n",
" Middle \n",
" Never \n",
" Employed \n",
" 0 \n",
" 0 \n",
" 0 \n",
" \n",
" \n",
" 299998 \n",
" 999998 \n",
" 48 \n",
" 3 \n",
" 72 \n",
" 4.9 \n",
" 6.9 \n",
" 1.8 \n",
" 27.7 \n",
" 0.89 \n",
" 121 \n",
" 67 \n",
" 86 \n",
" 171 \n",
" 50 \n",
" 84 \n",
" 138 \n",
" Male \n",
" White \n",
" Highschool \n",
" Low \n",
" Current \n",
" Retired \n",
" 0 \n",
" 1 \n",
" 0 \n",
" \n",
" \n",
" 299999 \n",
" 999999 \n",
" 47 \n",
" 1 \n",
" 75 \n",
" 5.1 \n",
" 8.0 \n",
" 2.5 \n",
" 26.1 \n",
" 0.84 \n",
" 132 \n",
" 70 \n",
" 87 \n",
" 193 \n",
" 46 \n",
" 109 \n",
" 107 \n",
" Female \n",
" White \n",
" Graduate \n",
" Upper-Middle \n",
" Current \n",
" Employed \n",
" 0 \n",
" 1 \n",
" 0 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id age alcohol_consumption_per_week \\\n",
"299995 999995 59 3 \n",
"299996 999996 50 2 \n",
"299997 999997 63 1 \n",
"299998 999998 48 3 \n",
"299999 999999 47 1 \n",
"\n",
" physical_activity_minutes_per_week diet_score sleep_hours_per_day \\\n",
"299995 185 6.3 7.3 \n",
"299996 25 5.8 7.8 \n",
"299997 252 5.2 7.5 \n",
"299998 72 4.9 6.9 \n",
"299999 75 5.1 8.0 \n",
"\n",
" screen_time_hours_per_day bmi waist_to_hip_ratio systolic_bp \\\n",
"299995 4.4 22.8 0.81 108 \n",
"299996 4.5 29.6 0.93 112 \n",
"299997 8.5 25.1 0.77 129 \n",
"299998 1.8 27.7 0.89 121 \n",
"299999 2.5 26.1 0.84 132 \n",
"\n",
" diastolic_bp heart_rate cholesterol_total hdl_cholesterol \\\n",
"299995 75 64 189 54 \n",
"299996 70 78 158 43 \n",
"299997 63 57 168 33 \n",
"299998 67 86 171 50 \n",
"299999 70 87 193 46 \n",
"\n",
" ldl_cholesterol triglycerides gender ethnicity education_level \\\n",
"299995 96 126 Male White Highschool \n",
"299996 84 112 Male Asian Postgraduate \n",
"299997 105 135 Female White Highschool \n",
"299998 84 138 Male White Highschool \n",
"299999 109 107 Female White Graduate \n",
"\n",
" income_level smoking_status employment_status \\\n",
"299995 Upper-Middle Former Employed \n",
"299996 Lower-Middle Never Employed \n",
"299997 Middle Never Employed \n",
"299998 Low Current Retired \n",
"299999 Upper-Middle Current Employed \n",
"\n",
" family_history_diabetes hypertension_history cardiovascular_history \n",
"299995 1 0 0 \n",
"299996 0 0 0 \n",
"299997 0 0 0 \n",
"299998 0 1 0 \n",
"299999 0 1 0 "
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df2.tail()"
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "3925eebe-064f-41d7-852a-86d752d3d03c",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" count \n",
" mean \n",
" std \n",
" min \n",
" 25% \n",
" 50% \n",
" 75% \n",
" max \n",
" \n",
" \n",
" \n",
" \n",
" id \n",
" 700000.0 \n",
" 349999.500000 \n",
" 202072.738554 \n",
" 0.00 \n",
" 174999.75 \n",
" 349999.50 \n",
" 524999.25 \n",
" 699999.00 \n",
" \n",
" \n",
" age \n",
" 700000.0 \n",
" 50.359734 \n",
" 11.655520 \n",
" 19.00 \n",
" 42.00 \n",
" 50.00 \n",
" 58.00 \n",
" 89.00 \n",
" \n",
" \n",
" alcohol_consumption_per_week \n",
" 700000.0 \n",
" 2.072411 \n",
" 1.048189 \n",
" 1.00 \n",
" 1.00 \n",
" 2.00 \n",
" 3.00 \n",
" 9.00 \n",
" \n",
" \n",
" physical_activity_minutes_per_week \n",
" 700000.0 \n",
" 80.230803 \n",
" 51.195071 \n",
" 1.00 \n",
" 49.00 \n",
" 71.00 \n",
" 96.00 \n",
" 747.00 \n",
" \n",
" \n",
" diet_score \n",
" 700000.0 \n",
" 5.963695 \n",
" 1.463336 \n",
" 0.10 \n",
" 5.00 \n",
" 6.00 \n",
" 7.00 \n",
" 9.90 \n",
" \n",
" \n",
" sleep_hours_per_day \n",
" 700000.0 \n",
" 7.002200 \n",
" 0.901907 \n",
" 3.10 \n",
" 6.40 \n",
" 7.00 \n",
" 7.60 \n",
" 9.90 \n",
" \n",
" \n",
" screen_time_hours_per_day \n",
" 700000.0 \n",
" 6.012733 \n",
" 2.022707 \n",
" 0.60 \n",
" 4.60 \n",
" 6.00 \n",
" 7.40 \n",
" 16.50 \n",
" \n",
" \n",
" bmi \n",
" 700000.0 \n",
" 25.874684 \n",
" 2.860705 \n",
" 15.10 \n",
" 23.90 \n",
" 25.90 \n",
" 27.80 \n",
" 38.40 \n",
" \n",
" \n",
" waist_to_hip_ratio \n",
" 700000.0 \n",
" 0.858766 \n",
" 0.037980 \n",
" 0.68 \n",
" 0.83 \n",
" 0.86 \n",
" 0.88 \n",
" 1.05 \n",
" \n",
" \n",
" systolic_bp \n",
" 700000.0 \n",
" 116.294193 \n",
" 11.010390 \n",
" 91.00 \n",
" 108.00 \n",
" 116.00 \n",
" 124.00 \n",
" 163.00 \n",
" \n",
" \n",
" diastolic_bp \n",
" 700000.0 \n",
" 75.440924 \n",
" 6.825775 \n",
" 51.00 \n",
" 71.00 \n",
" 75.00 \n",
" 80.00 \n",
" 104.00 \n",
" \n",
" \n",
" heart_rate \n",
" 700000.0 \n",
" 70.167749 \n",
" 6.938722 \n",
" 42.00 \n",
" 65.00 \n",
" 70.00 \n",
" 75.00 \n",
" 101.00 \n",
" \n",
" \n",
" cholesterol_total \n",
" 700000.0 \n",
" 186.818801 \n",
" 16.730832 \n",
" 117.00 \n",
" 175.00 \n",
" 187.00 \n",
" 199.00 \n",
" 289.00 \n",
" \n",
" \n",
" hdl_cholesterol \n",
" 700000.0 \n",
" 53.823214 \n",
" 8.266545 \n",
" 21.00 \n",
" 48.00 \n",
" 54.00 \n",
" 59.00 \n",
" 90.00 \n",
" \n",
" \n",
" ldl_cholesterol \n",
" 700000.0 \n",
" 102.905854 \n",
" 19.022416 \n",
" 51.00 \n",
" 89.00 \n",
" 103.00 \n",
" 116.00 \n",
" 205.00 \n",
" \n",
" \n",
" triglycerides \n",
" 700000.0 \n",
" 123.081850 \n",
" 24.739397 \n",
" 31.00 \n",
" 106.00 \n",
" 123.00 \n",
" 139.00 \n",
" 290.00 \n",
" \n",
" \n",
" family_history_diabetes \n",
" 700000.0 \n",
" 0.149401 \n",
" 0.356484 \n",
" 0.00 \n",
" 0.00 \n",
" 0.00 \n",
" 0.00 \n",
" 1.00 \n",
" \n",
" \n",
" hypertension_history \n",
" 700000.0 \n",
" 0.181990 \n",
" 0.385837 \n",
" 0.00 \n",
" 0.00 \n",
" 0.00 \n",
" 0.00 \n",
" 1.00 \n",
" \n",
" \n",
" cardiovascular_history \n",
" 700000.0 \n",
" 0.030324 \n",
" 0.171478 \n",
" 0.00 \n",
" 0.00 \n",
" 0.00 \n",
" 0.00 \n",
" 1.00 \n",
" \n",
" \n",
" diagnosed_diabetes \n",
" 700000.0 \n",
" 0.623296 \n",
" 0.484560 \n",
" 0.00 \n",
" 0.00 \n",
" 1.00 \n",
" 1.00 \n",
" 1.00 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" count mean std \\\n",
"id 700000.0 349999.500000 202072.738554 \n",
"age 700000.0 50.359734 11.655520 \n",
"alcohol_consumption_per_week 700000.0 2.072411 1.048189 \n",
"physical_activity_minutes_per_week 700000.0 80.230803 51.195071 \n",
"diet_score 700000.0 5.963695 1.463336 \n",
"sleep_hours_per_day 700000.0 7.002200 0.901907 \n",
"screen_time_hours_per_day 700000.0 6.012733 2.022707 \n",
"bmi 700000.0 25.874684 2.860705 \n",
"waist_to_hip_ratio 700000.0 0.858766 0.037980 \n",
"systolic_bp 700000.0 116.294193 11.010390 \n",
"diastolic_bp 700000.0 75.440924 6.825775 \n",
"heart_rate 700000.0 70.167749 6.938722 \n",
"cholesterol_total 700000.0 186.818801 16.730832 \n",
"hdl_cholesterol 700000.0 53.823214 8.266545 \n",
"ldl_cholesterol 700000.0 102.905854 19.022416 \n",
"triglycerides 700000.0 123.081850 24.739397 \n",
"family_history_diabetes 700000.0 0.149401 0.356484 \n",
"hypertension_history 700000.0 0.181990 0.385837 \n",
"cardiovascular_history 700000.0 0.030324 0.171478 \n",
"diagnosed_diabetes 700000.0 0.623296 0.484560 \n",
"\n",
" min 25% 50% 75% \\\n",
"id 0.00 174999.75 349999.50 524999.25 \n",
"age 19.00 42.00 50.00 58.00 \n",
"alcohol_consumption_per_week 1.00 1.00 2.00 3.00 \n",
"physical_activity_minutes_per_week 1.00 49.00 71.00 96.00 \n",
"diet_score 0.10 5.00 6.00 7.00 \n",
"sleep_hours_per_day 3.10 6.40 7.00 7.60 \n",
"screen_time_hours_per_day 0.60 4.60 6.00 7.40 \n",
"bmi 15.10 23.90 25.90 27.80 \n",
"waist_to_hip_ratio 0.68 0.83 0.86 0.88 \n",
"systolic_bp 91.00 108.00 116.00 124.00 \n",
"diastolic_bp 51.00 71.00 75.00 80.00 \n",
"heart_rate 42.00 65.00 70.00 75.00 \n",
"cholesterol_total 117.00 175.00 187.00 199.00 \n",
"hdl_cholesterol 21.00 48.00 54.00 59.00 \n",
"ldl_cholesterol 51.00 89.00 103.00 116.00 \n",
"triglycerides 31.00 106.00 123.00 139.00 \n",
"family_history_diabetes 0.00 0.00 0.00 0.00 \n",
"hypertension_history 0.00 0.00 0.00 0.00 \n",
"cardiovascular_history 0.00 0.00 0.00 0.00 \n",
"diagnosed_diabetes 0.00 0.00 1.00 1.00 \n",
"\n",
" max \n",
"id 699999.00 \n",
"age 89.00 \n",
"alcohol_consumption_per_week 9.00 \n",
"physical_activity_minutes_per_week 747.00 \n",
"diet_score 9.90 \n",
"sleep_hours_per_day 9.90 \n",
"screen_time_hours_per_day 16.50 \n",
"bmi 38.40 \n",
"waist_to_hip_ratio 1.05 \n",
"systolic_bp 163.00 \n",
"diastolic_bp 104.00 \n",
"heart_rate 101.00 \n",
"cholesterol_total 289.00 \n",
"hdl_cholesterol 90.00 \n",
"ldl_cholesterol 205.00 \n",
"triglycerides 290.00 \n",
"family_history_diabetes 1.00 \n",
"hypertension_history 1.00 \n",
"cardiovascular_history 1.00 \n",
"diagnosed_diabetes 1.00 "
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df1.describe().T"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "e787411d-c670-44dc-86a0-1a6bdc3679db",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"RangeIndex: 700000 entries, 0 to 699999\n",
"Data columns (total 26 columns):\n",
" # Column Non-Null Count Dtype \n",
"--- ------ -------------- ----- \n",
" 0 id 700000 non-null int64 \n",
" 1 age 700000 non-null int64 \n",
" 2 alcohol_consumption_per_week 700000 non-null int64 \n",
" 3 physical_activity_minutes_per_week 700000 non-null int64 \n",
" 4 diet_score 700000 non-null float64\n",
" 5 sleep_hours_per_day 700000 non-null float64\n",
" 6 screen_time_hours_per_day 700000 non-null float64\n",
" 7 bmi 700000 non-null float64\n",
" 8 waist_to_hip_ratio 700000 non-null float64\n",
" 9 systolic_bp 700000 non-null int64 \n",
" 10 diastolic_bp 700000 non-null int64 \n",
" 11 heart_rate 700000 non-null int64 \n",
" 12 cholesterol_total 700000 non-null int64 \n",
" 13 hdl_cholesterol 700000 non-null int64 \n",
" 14 ldl_cholesterol 700000 non-null int64 \n",
" 15 triglycerides 700000 non-null int64 \n",
" 16 gender 700000 non-null object \n",
" 17 ethnicity 700000 non-null object \n",
" 18 education_level 700000 non-null object \n",
" 19 income_level 700000 non-null object \n",
" 20 smoking_status 700000 non-null object \n",
" 21 employment_status 700000 non-null object \n",
" 22 family_history_diabetes 700000 non-null int64 \n",
" 23 hypertension_history 700000 non-null int64 \n",
" 24 cardiovascular_history 700000 non-null int64 \n",
" 25 diagnosed_diabetes 700000 non-null float64\n",
"dtypes: float64(6), int64(14), object(6)\n",
"memory usage: 138.9+ MB\n"
]
}
],
"source": [
"df1.info()"
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "184e4ea2-64ad-4bb2-8184-32e26b24ae0d",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"RangeIndex: 300000 entries, 0 to 299999\n",
"Data columns (total 25 columns):\n",
" # Column Non-Null Count Dtype \n",
"--- ------ -------------- ----- \n",
" 0 id 300000 non-null int64 \n",
" 1 age 300000 non-null int64 \n",
" 2 alcohol_consumption_per_week 300000 non-null int64 \n",
" 3 physical_activity_minutes_per_week 300000 non-null int64 \n",
" 4 diet_score 300000 non-null float64\n",
" 5 sleep_hours_per_day 300000 non-null float64\n",
" 6 screen_time_hours_per_day 300000 non-null float64\n",
" 7 bmi 300000 non-null float64\n",
" 8 waist_to_hip_ratio 300000 non-null float64\n",
" 9 systolic_bp 300000 non-null int64 \n",
" 10 diastolic_bp 300000 non-null int64 \n",
" 11 heart_rate 300000 non-null int64 \n",
" 12 cholesterol_total 300000 non-null int64 \n",
" 13 hdl_cholesterol 300000 non-null int64 \n",
" 14 ldl_cholesterol 300000 non-null int64 \n",
" 15 triglycerides 300000 non-null int64 \n",
" 16 gender 300000 non-null object \n",
" 17 ethnicity 300000 non-null object \n",
" 18 education_level 300000 non-null object \n",
" 19 income_level 300000 non-null object \n",
" 20 smoking_status 300000 non-null object \n",
" 21 employment_status 300000 non-null object \n",
" 22 family_history_diabetes 300000 non-null int64 \n",
" 23 hypertension_history 300000 non-null int64 \n",
" 24 cardiovascular_history 300000 non-null int64 \n",
"dtypes: float64(5), int64(14), object(6)\n",
"memory usage: 57.2+ MB\n"
]
}
],
"source": [
"df2.info()"
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "f7556172-ac48-43fb-aff7-8e60a32e6cc2",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"id 0.0\n",
"age 0.0\n",
"alcohol_consumption_per_week 0.0\n",
"physical_activity_minutes_per_week 0.0\n",
"diet_score 0.0\n",
"sleep_hours_per_day 0.0\n",
"screen_time_hours_per_day 0.0\n",
"bmi 0.0\n",
"waist_to_hip_ratio 0.0\n",
"systolic_bp 0.0\n",
"diastolic_bp 0.0\n",
"heart_rate 0.0\n",
"cholesterol_total 0.0\n",
"hdl_cholesterol 0.0\n",
"ldl_cholesterol 0.0\n",
"triglycerides 0.0\n",
"gender 0.0\n",
"ethnicity 0.0\n",
"education_level 0.0\n",
"income_level 0.0\n",
"smoking_status 0.0\n",
"employment_status 0.0\n",
"family_history_diabetes 0.0\n",
"hypertension_history 0.0\n",
"cardiovascular_history 0.0\n",
"diagnosed_diabetes 0.0\n",
"dtype: float64"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df1.isnull().mean()"
]
},
{
"cell_type": "code",
"execution_count": 13,
"id": "06753328-b281-4348-9966-7b4b403819ec",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"id 0.0\n",
"age 0.0\n",
"alcohol_consumption_per_week 0.0\n",
"physical_activity_minutes_per_week 0.0\n",
"diet_score 0.0\n",
"sleep_hours_per_day 0.0\n",
"screen_time_hours_per_day 0.0\n",
"bmi 0.0\n",
"waist_to_hip_ratio 0.0\n",
"systolic_bp 0.0\n",
"diastolic_bp 0.0\n",
"heart_rate 0.0\n",
"cholesterol_total 0.0\n",
"hdl_cholesterol 0.0\n",
"ldl_cholesterol 0.0\n",
"triglycerides 0.0\n",
"gender 0.0\n",
"ethnicity 0.0\n",
"education_level 0.0\n",
"income_level 0.0\n",
"smoking_status 0.0\n",
"employment_status 0.0\n",
"family_history_diabetes 0.0\n",
"hypertension_history 0.0\n",
"cardiovascular_history 0.0\n",
"dtype: float64"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df2.isnull().mean()"
]
},
{
"cell_type": "code",
"execution_count": 35,
"id": "66f7b1d2-3007-4fd5-93f8-23dfe7668db4",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(700000, 35)"
]
},
"execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df1.shape"
]
},
{
"cell_type": "code",
"execution_count": 36,
"id": "e8db92fd-fdf7-4aa0-8486-d69a91d82a33",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"(300000, 34)"
]
},
"execution_count": 36,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df2.shape"
]
},
{
"cell_type": "code",
"execution_count": 15,
"id": "cf88c27d-fbb4-4b94-a885-53fca5bd5ae4",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"plt.figure(figsize=(12,8))\n",
"sns.heatmap(df1.corr(numeric_only=True)[['diagnosed_diabetes']].sort_values('diagnosed_diabetes', ascending=False),annot=True, cmap='coolwarm')\n",
"plt.title('Correlation with Diagnosed Diabetes')\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"id": "7fa38157-5808-437e-9186-2e7cf163245b",
"metadata": {},
"source": [
"Family history of diabetes, age, BMI, and blood pressure show the strongest positive correlation with diabetes, while physical activity is negatively correlated."
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "05783e50-1b61-410b-a315-65aacd08b491",
"metadata": {},
"outputs": [],
"source": [
"df=pd.concat([df1,df2])"
]
},
{
"cell_type": "code",
"execution_count": 16,
"id": "a0f4e8b1-0366-47fa-b81d-570318869f5f",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"diagnosed_diabetes\n",
"1.0 0.623296\n",
"0.0 0.376704\n",
"Name: proportion, dtype: float64"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df1['diagnosed_diabetes'].value_counts(normalize=True)"
]
},
{
"cell_type": "code",
"execution_count": 19,
"id": "c9def00d-86e5-488d-a627-9a4be681ef7e",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"sns.countplot(data=df1, x='diagnosed_diabetes');"
]
},
{
"cell_type": "markdown",
"id": "488a3c3e-f097-40c3-ae49-41cedc04db8f",
"metadata": {},
"source": [
"The dataset shows a moderate class imbalance, with more positive diabetes cases than negative ones."
]
},
{
"cell_type": "code",
"execution_count": 26,
"id": "0cff041c-a960-4c5b-b14b-ad96e6d1736b",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"sns.boxplot(data=df1,x='diagnosed_diabetes',y='bmi');"
]
},
{
"cell_type": "markdown",
"id": "b355f2a7-6e02-4a87-b783-ccfc66f2feb7",
"metadata": {},
"source": [
"Patients diagnosed with diabetes tend to have a higher median BMI compared to non-diabetic patients."
]
},
{
"cell_type": "code",
"execution_count": 27,
"id": "88d144dc-b1d7-444c-9255-6f72712e26e5",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"sns.boxplot(data=df1,x='diagnosed_diabetes',y='age');"
]
},
{
"cell_type": "markdown",
"id": "a9255e3d-3f48-48c1-bf5b-36633f27525c",
"metadata": {},
"source": [
"Individuals with diabetes are generally older, with a higher median age than non-diabetic individuals."
]
},
{
"cell_type": "code",
"execution_count": 23,
"id": "4d07fadf-3dfe-4453-89ac-b07dda3cb043",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"sns.boxplot(data=df,y='bmi');"
]
},
{
"cell_type": "markdown",
"id": "5f35775e-93a3-4501-8d7b-afea4d4b359c",
"metadata": {},
"source": [
"BMI values are mostly concentrated in the overweight range, with some high but still realistic outliers."
]
},
{
"cell_type": "code",
"execution_count": 29,
"id": "570670ff-55df-49e6-814c-0b531e24de1b",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"gender\n",
"Female 517335\n",
"Male 477187\n",
"Other 5478\n",
"Name: count, dtype: int64"
]
},
"execution_count": 29,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df['gender'].value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 30,
"id": "02269f6a-b7d5-4f9d-99ed-c91924f6eabe",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"ethnicity\n",
"White 554528\n",
"Hispanic 183518\n",
"Black 151504\n",
"Asian 85197\n",
"Other 25253\n",
"Name: count, dtype: int64"
]
},
"execution_count": 30,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df['ethnicity'].value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 31,
"id": "de33fa09-ad81-48ac-8fb8-f625d78916e4",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"education_level\n",
"Highschool 497500\n",
"Graduate 366708\n",
"Postgraduate 113760\n",
"No formal 22032\n",
"Name: count, dtype: int64"
]
},
"execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df['education_level'].value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 32,
"id": "29c0dd9e-e115-49f6-89cf-ea7a9c9e167c",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"income_level\n",
"Middle 414806\n",
"Lower-Middle 254855\n",
"Upper-Middle 182460\n",
"Low 122619\n",
"High 25260\n",
"Name: count, dtype: int64"
]
},
"execution_count": 32,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df['income_level'].value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 33,
"id": "0685d86d-271e-480b-b4fa-0ccc74deca2e",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"smoking_status\n",
"Never 706114\n",
"Current 147259\n",
"Former 146627\n",
"Name: count, dtype: int64"
]
},
"execution_count": 33,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df['smoking_status'].value_counts()"
]
},
{
"cell_type": "code",
"execution_count": 34,
"id": "3e24b2e7-ff81-49ce-bcd0-46ba163f891e",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"employment_status\n",
"Employed 734163\n",
"Retired 166942\n",
"Unemployed 71668\n",
"Student 27227\n",
"Name: count, dtype: int64"
]
},
"execution_count": 34,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df['employment_status'].value_counts()"
]
},
{
"cell_type": "markdown",
"id": "072c204b-cd27-4aa6-b0f1-e6411af580eb",
"metadata": {},
"source": [
"### Feature Engineering"
]
},
{
"cell_type": "code",
"execution_count": 41,
"id": "40376953-a6cb-47bf-869f-c30cf602726a",
"metadata": {},
"outputs": [],
"source": [
"eps=1e-6\n",
"# Body / Age\n",
"df['bmi_age_ratio']=df['bmi'] / (df['age'] + eps)\n",
"df['is_obese']=(df['bmi'] >= 30).astype(int)\n",
"\n",
"# Activity features\n",
"df['activity_per_age']=(df['physical_activity_minutes_per_week'] / (df['age'] + eps))\n",
"df['low_activity']=(df['physical_activity_minutes_per_week'] < 150).astype(int)\n",
"\n",
"# Sleep features \n",
"df['sleep_deficit']=7 - df['sleep_hours_per_day']\n",
"df['poor_sleep']=(df['sleep_hours_per_day'] < 6).astype(int)\n",
"\n",
"# Cholesterol ?\n",
"if {'ldl_cholesterol', 'hdl_cholesterol'}.issubset(df.columns):\n",
" df['chol_ratio']=(df['ldl_cholesterol'] / (df['hdl_cholesterol'] + eps))\n",
" df['high_ldl']=(df['ldl_cholesterol'] >= 130).astype(int)\n",
" df['low_hdl']=(df['hdl_cholesterol'] < 40).astype(int)"
]
},
{
"cell_type": "markdown",
"id": "d2de5e58-ff81-4687-b996-7bda4f5f3680",
"metadata": {},
"source": [
"Feature engineering creates age-adjusted body, activity, sleep, and cholesterol indicators using rule-based transformations that are meaningful."
]
},
{
"cell_type": "code",
"execution_count": 44,
"id": "9279601c-5b1a-47f1-bed4-d27d619f6b93",
"metadata": {},
"outputs": [],
"source": [
"df=pd.get_dummies(df,drop_first=True)"
]
},
{
"cell_type": "code",
"execution_count": 45,
"id": "60eafedc-cec6-4aaf-8a78-750f731c3f06",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Index: 1000000 entries, 0 to 299999\n",
"Data columns (total 47 columns):\n",
" # Column Non-Null Count Dtype \n",
"--- ------ -------------- ----- \n",
" 0 id 1000000 non-null int64 \n",
" 1 age 1000000 non-null int64 \n",
" 2 alcohol_consumption_per_week 1000000 non-null int64 \n",
" 3 physical_activity_minutes_per_week 1000000 non-null int64 \n",
" 4 diet_score 1000000 non-null float64\n",
" 5 sleep_hours_per_day 1000000 non-null float64\n",
" 6 screen_time_hours_per_day 1000000 non-null float64\n",
" 7 bmi 1000000 non-null float64\n",
" 8 waist_to_hip_ratio 1000000 non-null float64\n",
" 9 systolic_bp 1000000 non-null int64 \n",
" 10 diastolic_bp 1000000 non-null int64 \n",
" 11 heart_rate 1000000 non-null int64 \n",
" 12 cholesterol_total 1000000 non-null int64 \n",
" 13 hdl_cholesterol 1000000 non-null int64 \n",
" 14 ldl_cholesterol 1000000 non-null int64 \n",
" 15 triglycerides 1000000 non-null int64 \n",
" 16 family_history_diabetes 1000000 non-null int64 \n",
" 17 hypertension_history 1000000 non-null int64 \n",
" 18 cardiovascular_history 1000000 non-null int64 \n",
" 19 diagnosed_diabetes 700000 non-null float64\n",
" 20 bmi_age_ratio 1000000 non-null float64\n",
" 21 is_obese 1000000 non-null int64 \n",
" 22 activity_per_age 1000000 non-null float64\n",
" 23 low_activity 1000000 non-null int64 \n",
" 24 sleep_deficit 1000000 non-null float64\n",
" 25 poor_sleep 1000000 non-null int64 \n",
" 26 chol_ratio 1000000 non-null float64\n",
" 27 high_ldl 1000000 non-null int64 \n",
" 28 low_hdl 1000000 non-null int64 \n",
" 29 gender_Male 1000000 non-null bool \n",
" 30 gender_Other 1000000 non-null bool \n",
" 31 ethnicity_Black 1000000 non-null bool \n",
" 32 ethnicity_Hispanic 1000000 non-null bool \n",
" 33 ethnicity_Other 1000000 non-null bool \n",
" 34 ethnicity_White 1000000 non-null bool \n",
" 35 education_level_Highschool 1000000 non-null bool \n",
" 36 education_level_No formal 1000000 non-null bool \n",
" 37 education_level_Postgraduate 1000000 non-null bool \n",
" 38 income_level_Low 1000000 non-null bool \n",
" 39 income_level_Lower-Middle 1000000 non-null bool \n",
" 40 income_level_Middle 1000000 non-null bool \n",
" 41 income_level_Upper-Middle 1000000 non-null bool \n",
" 42 smoking_status_Former 1000000 non-null bool \n",
" 43 smoking_status_Never 1000000 non-null bool \n",
" 44 employment_status_Retired 1000000 non-null bool \n",
" 45 employment_status_Student 1000000 non-null bool \n",
" 46 employment_status_Unemployed 1000000 non-null bool \n",
"dtypes: bool(18), float64(10), int64(19)\n",
"memory usage: 246.0 MB\n"
]
}
],
"source": [
"df.info()"
]
},
{
"cell_type": "markdown",
"id": "bdb3ceff-1dc9-493c-a790-1fc06f6e2093",
"metadata": {},
"source": [
"### Modelling"
]
},
{
"cell_type": "code",
"execution_count": 47,
"id": "4152ea8f-72cd-49e3-b374-422d193f3f63",
"metadata": {},
"outputs": [],
"source": [
"train=df[:700000]\n",
"test=df[700000:]"
]
},
{
"cell_type": "code",
"execution_count": 48,
"id": "7d75e6c5-03e8-4228-84b8-471ab55d5f4d",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" age \n",
" alcohol_consumption_per_week \n",
" physical_activity_minutes_per_week \n",
" diet_score \n",
" sleep_hours_per_day \n",
" screen_time_hours_per_day \n",
" bmi \n",
" waist_to_hip_ratio \n",
" systolic_bp \n",
" diastolic_bp \n",
" heart_rate \n",
" cholesterol_total \n",
" hdl_cholesterol \n",
" ldl_cholesterol \n",
" triglycerides \n",
" family_history_diabetes \n",
" hypertension_history \n",
" cardiovascular_history \n",
" diagnosed_diabetes \n",
" bmi_age_ratio \n",
" is_obese \n",
" activity_per_age \n",
" low_activity \n",
" sleep_deficit \n",
" poor_sleep \n",
" chol_ratio \n",
" high_ldl \n",
" low_hdl \n",
" gender_Male \n",
" gender_Other \n",
" ethnicity_Black \n",
" ethnicity_Hispanic \n",
" ethnicity_Other \n",
" ethnicity_White \n",
" education_level_Highschool \n",
" education_level_No formal \n",
" education_level_Postgraduate \n",
" income_level_Low \n",
" income_level_Lower-Middle \n",
" income_level_Middle \n",
" income_level_Upper-Middle \n",
" smoking_status_Former \n",
" smoking_status_Never \n",
" employment_status_Retired \n",
" employment_status_Student \n",
" employment_status_Unemployed \n",
" \n",
" \n",
" \n",
" \n",
" 699995 \n",
" 699995 \n",
" 29 \n",
" 1 \n",
" 59 \n",
" 6.9 \n",
" 5.2 \n",
" 1.5 \n",
" 26.1 \n",
" 0.88 \n",
" 133 \n",
" 57 \n",
" 69 \n",
" 163 \n",
" 58 \n",
" 90 \n",
" 126 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 0.0 \n",
" 0.900000 \n",
" 0 \n",
" 2.034483 \n",
" 1 \n",
" 1.8 \n",
" 1 \n",
" 1.551724 \n",
" 0 \n",
" 0 \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" True \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" \n",
" \n",
" 699996 \n",
" 699996 \n",
" 46 \n",
" 2 \n",
" 72 \n",
" 7.7 \n",
" 7.7 \n",
" 3.8 \n",
" 25.5 \n",
" 0.85 \n",
" 106 \n",
" 85 \n",
" 65 \n",
" 188 \n",
" 45 \n",
" 107 \n",
" 119 \n",
" 0 \n",
" 0 \n",
" 1 \n",
" 1.0 \n",
" 0.554348 \n",
" 0 \n",
" 1.565217 \n",
" 1 \n",
" -0.7 \n",
" 0 \n",
" 2.377778 \n",
" 0 \n",
" 0 \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" \n",
" \n",
" 699997 \n",
" 699997 \n",
" 35 \n",
" 1 \n",
" 50 \n",
" 5.6 \n",
" 6.1 \n",
" 6.4 \n",
" 26.9 \n",
" 0.88 \n",
" 127 \n",
" 84 \n",
" 63 \n",
" 168 \n",
" 59 \n",
" 77 \n",
" 166 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" 0.768571 \n",
" 0 \n",
" 1.428571 \n",
" 1 \n",
" 0.9 \n",
" 0 \n",
" 1.305085 \n",
" 0 \n",
" 0 \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" \n",
" \n",
" 699998 \n",
" 699998 \n",
" 49 \n",
" 2 \n",
" 70 \n",
" 5.7 \n",
" 6.9 \n",
" 4.7 \n",
" 25.2 \n",
" 0.86 \n",
" 116 \n",
" 67 \n",
" 69 \n",
" 198 \n",
" 55 \n",
" 108 \n",
" 133 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" 0.514286 \n",
" 0 \n",
" 1.428571 \n",
" 1 \n",
" 0.1 \n",
" 0 \n",
" 1.963636 \n",
" 0 \n",
" 0 \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" True \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" True \n",
" True \n",
" False \n",
" False \n",
" \n",
" \n",
" 699999 \n",
" 699999 \n",
" 42 \n",
" 4 \n",
" 56 \n",
" 4.2 \n",
" 7.0 \n",
" 6.1 \n",
" 28.3 \n",
" 0.90 \n",
" 99 \n",
" 73 \n",
" 74 \n",
" 191 \n",
" 63 \n",
" 105 \n",
" 159 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" 1.0 \n",
" 0.673810 \n",
" 0 \n",
" 1.333333 \n",
" 1 \n",
" 0.0 \n",
" 0 \n",
" 1.666667 \n",
" 0 \n",
" 0 \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id age alcohol_consumption_per_week \\\n",
"699995 699995 29 1 \n",
"699996 699996 46 2 \n",
"699997 699997 35 1 \n",
"699998 699998 49 2 \n",
"699999 699999 42 4 \n",
"\n",
" physical_activity_minutes_per_week diet_score sleep_hours_per_day \\\n",
"699995 59 6.9 5.2 \n",
"699996 72 7.7 7.7 \n",
"699997 50 5.6 6.1 \n",
"699998 70 5.7 6.9 \n",
"699999 56 4.2 7.0 \n",
"\n",
" screen_time_hours_per_day bmi waist_to_hip_ratio systolic_bp \\\n",
"699995 1.5 26.1 0.88 133 \n",
"699996 3.8 25.5 0.85 106 \n",
"699997 6.4 26.9 0.88 127 \n",
"699998 4.7 25.2 0.86 116 \n",
"699999 6.1 28.3 0.90 99 \n",
"\n",
" diastolic_bp heart_rate cholesterol_total hdl_cholesterol \\\n",
"699995 57 69 163 58 \n",
"699996 85 65 188 45 \n",
"699997 84 63 168 59 \n",
"699998 67 69 198 55 \n",
"699999 73 74 191 63 \n",
"\n",
" ldl_cholesterol triglycerides family_history_diabetes \\\n",
"699995 90 126 0 \n",
"699996 107 119 0 \n",
"699997 77 166 0 \n",
"699998 108 133 0 \n",
"699999 105 159 0 \n",
"\n",
" hypertension_history cardiovascular_history diagnosed_diabetes \\\n",
"699995 0 0 0.0 \n",
"699996 0 1 1.0 \n",
"699997 0 0 1.0 \n",
"699998 0 0 1.0 \n",
"699999 0 0 1.0 \n",
"\n",
" bmi_age_ratio is_obese activity_per_age low_activity \\\n",
"699995 0.900000 0 2.034483 1 \n",
"699996 0.554348 0 1.565217 1 \n",
"699997 0.768571 0 1.428571 1 \n",
"699998 0.514286 0 1.428571 1 \n",
"699999 0.673810 0 1.333333 1 \n",
"\n",
" sleep_deficit poor_sleep chol_ratio high_ldl low_hdl gender_Male \\\n",
"699995 1.8 1 1.551724 0 0 False \n",
"699996 -0.7 0 2.377778 0 0 False \n",
"699997 0.9 0 1.305085 0 0 False \n",
"699998 0.1 0 1.963636 0 0 False \n",
"699999 0.0 0 1.666667 0 0 False \n",
"\n",
" gender_Other ethnicity_Black ethnicity_Hispanic ethnicity_Other \\\n",
"699995 False False True False \n",
"699996 False False True False \n",
"699997 False False False False \n",
"699998 False False False False \n",
"699999 False True False False \n",
"\n",
" ethnicity_White education_level_Highschool \\\n",
"699995 False False \n",
"699996 False False \n",
"699997 True False \n",
"699998 True True \n",
"699999 False False \n",
"\n",
" education_level_No formal education_level_Postgraduate \\\n",
"699995 False True \n",
"699996 False False \n",
"699997 False False \n",
"699998 False False \n",
"699999 False False \n",
"\n",
" income_level_Low income_level_Lower-Middle income_level_Middle \\\n",
"699995 False False False \n",
"699996 False False False \n",
"699997 False False True \n",
"699998 False True False \n",
"699999 True False False \n",
"\n",
" income_level_Upper-Middle smoking_status_Former \\\n",
"699995 True True \n",
"699996 True True \n",
"699997 False False \n",
"699998 False False \n",
"699999 False True \n",
"\n",
" smoking_status_Never employment_status_Retired \\\n",
"699995 False False \n",
"699996 False False \n",
"699997 True False \n",
"699998 True True \n",
"699999 False False \n",
"\n",
" employment_status_Student employment_status_Unemployed \n",
"699995 False False \n",
"699996 False False \n",
"699997 False False \n",
"699998 False False \n",
"699999 False False "
]
},
"execution_count": 48,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"train.tail()"
]
},
{
"cell_type": "code",
"execution_count": 49,
"id": "de54c51a-478f-49c5-b43b-bf1e1024f7ad",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" age \n",
" alcohol_consumption_per_week \n",
" physical_activity_minutes_per_week \n",
" diet_score \n",
" sleep_hours_per_day \n",
" screen_time_hours_per_day \n",
" bmi \n",
" waist_to_hip_ratio \n",
" systolic_bp \n",
" diastolic_bp \n",
" heart_rate \n",
" cholesterol_total \n",
" hdl_cholesterol \n",
" ldl_cholesterol \n",
" triglycerides \n",
" family_history_diabetes \n",
" hypertension_history \n",
" cardiovascular_history \n",
" diagnosed_diabetes \n",
" bmi_age_ratio \n",
" is_obese \n",
" activity_per_age \n",
" low_activity \n",
" sleep_deficit \n",
" poor_sleep \n",
" chol_ratio \n",
" high_ldl \n",
" low_hdl \n",
" gender_Male \n",
" gender_Other \n",
" ethnicity_Black \n",
" ethnicity_Hispanic \n",
" ethnicity_Other \n",
" ethnicity_White \n",
" education_level_Highschool \n",
" education_level_No formal \n",
" education_level_Postgraduate \n",
" income_level_Low \n",
" income_level_Lower-Middle \n",
" income_level_Middle \n",
" income_level_Upper-Middle \n",
" smoking_status_Former \n",
" smoking_status_Never \n",
" employment_status_Retired \n",
" employment_status_Student \n",
" employment_status_Unemployed \n",
" \n",
" \n",
" \n",
" \n",
" 0 \n",
" 700000 \n",
" 45 \n",
" 4 \n",
" 100 \n",
" 4.3 \n",
" 6.8 \n",
" 6.2 \n",
" 25.5 \n",
" 0.84 \n",
" 123 \n",
" 70 \n",
" 64 \n",
" 209 \n",
" 55 \n",
" 135 \n",
" 111 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" NaN \n",
" 0.566667 \n",
" 0 \n",
" 2.222222 \n",
" 1 \n",
" 0.2 \n",
" 0 \n",
" 2.454545 \n",
" 1 \n",
" 0 \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" \n",
" \n",
" 1 \n",
" 700001 \n",
" 35 \n",
" 1 \n",
" 87 \n",
" 3.5 \n",
" 4.6 \n",
" 9.0 \n",
" 28.6 \n",
" 0.88 \n",
" 120 \n",
" 74 \n",
" 59 \n",
" 159 \n",
" 47 \n",
" 83 \n",
" 145 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" NaN \n",
" 0.817143 \n",
" 0 \n",
" 2.485714 \n",
" 1 \n",
" 2.4 \n",
" 1 \n",
" 1.765957 \n",
" 0 \n",
" 0 \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" True \n",
" \n",
" \n",
" 2 \n",
" 700002 \n",
" 45 \n",
" 1 \n",
" 61 \n",
" 7.6 \n",
" 6.8 \n",
" 7.0 \n",
" 28.5 \n",
" 0.94 \n",
" 112 \n",
" 71 \n",
" 75 \n",
" 173 \n",
" 43 \n",
" 99 \n",
" 184 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" NaN \n",
" 0.633333 \n",
" 0 \n",
" 1.355556 \n",
" 1 \n",
" 0.2 \n",
" 0 \n",
" 2.302326 \n",
" 0 \n",
" 0 \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" True \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" \n",
" \n",
" 3 \n",
" 700003 \n",
" 55 \n",
" 2 \n",
" 81 \n",
" 7.3 \n",
" 7.3 \n",
" 5.0 \n",
" 26.9 \n",
" 0.91 \n",
" 114 \n",
" 81 \n",
" 61 \n",
" 203 \n",
" 59 \n",
" 116 \n",
" 128 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" NaN \n",
" 0.489091 \n",
" 0 \n",
" 1.472727 \n",
" 1 \n",
" -0.3 \n",
" 0 \n",
" 1.966102 \n",
" 0 \n",
" 0 \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" \n",
" \n",
" 4 \n",
" 700004 \n",
" 77 \n",
" 2 \n",
" 29 \n",
" 7.3 \n",
" 7.6 \n",
" 8.5 \n",
" 22.0 \n",
" 0.83 \n",
" 131 \n",
" 78 \n",
" 79 \n",
" 177 \n",
" 59 \n",
" 87 \n",
" 133 \n",
" 0 \n",
" 0 \n",
" 0 \n",
" NaN \n",
" 0.285714 \n",
" 0 \n",
" 0.376623 \n",
" 1 \n",
" -0.6 \n",
" 0 \n",
" 1.474576 \n",
" 0 \n",
" 0 \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" True \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" False \n",
" True \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id age alcohol_consumption_per_week \\\n",
"0 700000 45 4 \n",
"1 700001 35 1 \n",
"2 700002 45 1 \n",
"3 700003 55 2 \n",
"4 700004 77 2 \n",
"\n",
" physical_activity_minutes_per_week diet_score sleep_hours_per_day \\\n",
"0 100 4.3 6.8 \n",
"1 87 3.5 4.6 \n",
"2 61 7.6 6.8 \n",
"3 81 7.3 7.3 \n",
"4 29 7.3 7.6 \n",
"\n",
" screen_time_hours_per_day bmi waist_to_hip_ratio systolic_bp \\\n",
"0 6.2 25.5 0.84 123 \n",
"1 9.0 28.6 0.88 120 \n",
"2 7.0 28.5 0.94 112 \n",
"3 5.0 26.9 0.91 114 \n",
"4 8.5 22.0 0.83 131 \n",
"\n",
" diastolic_bp heart_rate cholesterol_total hdl_cholesterol \\\n",
"0 70 64 209 55 \n",
"1 74 59 159 47 \n",
"2 71 75 173 43 \n",
"3 81 61 203 59 \n",
"4 78 79 177 59 \n",
"\n",
" ldl_cholesterol triglycerides family_history_diabetes \\\n",
"0 135 111 0 \n",
"1 83 145 0 \n",
"2 99 184 0 \n",
"3 116 128 0 \n",
"4 87 133 0 \n",
"\n",
" hypertension_history cardiovascular_history diagnosed_diabetes \\\n",
"0 0 0 NaN \n",
"1 0 0 NaN \n",
"2 0 0 NaN \n",
"3 0 0 NaN \n",
"4 0 0 NaN \n",
"\n",
" bmi_age_ratio is_obese activity_per_age low_activity sleep_deficit \\\n",
"0 0.566667 0 2.222222 1 0.2 \n",
"1 0.817143 0 2.485714 1 2.4 \n",
"2 0.633333 0 1.355556 1 0.2 \n",
"3 0.489091 0 1.472727 1 -0.3 \n",
"4 0.285714 0 0.376623 1 -0.6 \n",
"\n",
" poor_sleep chol_ratio high_ldl low_hdl gender_Male gender_Other \\\n",
"0 0 2.454545 1 0 False False \n",
"1 1 1.765957 0 0 False False \n",
"2 0 2.302326 0 0 True False \n",
"3 0 1.966102 0 0 True False \n",
"4 0 1.474576 0 0 True False \n",
"\n",
" ethnicity_Black ethnicity_Hispanic ethnicity_Other ethnicity_White \\\n",
"0 False False False True \n",
"1 False False False True \n",
"2 False False False True \n",
"3 False False False True \n",
"4 False False False True \n",
"\n",
" education_level_Highschool education_level_No formal \\\n",
"0 True False \n",
"1 True False \n",
"2 True False \n",
"3 False False \n",
"4 False False \n",
"\n",
" education_level_Postgraduate income_level_Low income_level_Lower-Middle \\\n",
"0 False False False \n",
"1 False False False \n",
"2 False True False \n",
"3 False False False \n",
"4 False True False \n",
"\n",
" income_level_Middle income_level_Upper-Middle smoking_status_Former \\\n",
"0 True False True \n",
"1 True False False \n",
"2 False False False \n",
"3 True False True \n",
"4 False False False \n",
"\n",
" smoking_status_Never employment_status_Retired employment_status_Student \\\n",
"0 False False False \n",
"1 True False False \n",
"2 True False False \n",
"3 False False False \n",
"4 False False False \n",
"\n",
" employment_status_Unemployed \n",
"0 False \n",
"1 True \n",
"2 False \n",
"3 False \n",
"4 True "
]
},
"execution_count": 49,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"test.head()"
]
},
{
"cell_type": "code",
"execution_count": 51,
"id": "6168475c-0745-462e-8bda-02b50fa27cda",
"metadata": {},
"outputs": [],
"source": [
"x=train.drop(['diagnosed_diabetes', 'id'], axis=1)"
]
},
{
"cell_type": "code",
"execution_count": 52,
"id": "51deee8e-9c12-432c-9836-ed1a126f16dd",
"metadata": {},
"outputs": [],
"source": [
"y=train['diagnosed_diabetes']"
]
},
{
"cell_type": "code",
"execution_count": 53,
"id": "4d156cf7-6925-4ade-8d09-bcbd9f4eae98",
"metadata": {},
"outputs": [],
"source": [
"x_train,x_val, y_train, y_val=train_test_split(x,y, test_size=.20, random_state=42, stratify=y)"
]
},
{
"cell_type": "markdown",
"id": "68a8c1ac-4af9-4b7d-babc-6e91504bcbc6",
"metadata": {},
"source": [
"##### LogisticRegression"
]
},
{
"cell_type": "code",
"execution_count": 54,
"id": "626255ab-69c8-42c7-87c4-ecd1060edbb9",
"metadata": {},
"outputs": [],
"source": [
"lr=LogisticRegression(max_iter=1000,class_weight='balanced',n_jobs=-1)"
]
},
{
"cell_type": "code",
"execution_count": 55,
"id": "c861f536-bcdd-4c32-836c-03f92474866b",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"LogisticRegression(class_weight='balanced', max_iter=1000, n_jobs=-1) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. \n",
"
\n",
"
\n",
" Parameters \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" penalty \n",
" 'l2' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" dual \n",
" False \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" tol \n",
" 0.0001 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" C \n",
" 1.0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" fit_intercept \n",
" True \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" intercept_scaling \n",
" 1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" class_weight \n",
" 'balanced' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" random_state \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" solver \n",
" 'lbfgs' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_iter \n",
" 1000 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" multi_class \n",
" 'deprecated' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" verbose \n",
" 0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" warm_start \n",
" False \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" n_jobs \n",
" -1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" l1_ratio \n",
" None \n",
" \n",
" \n",
" \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
"LogisticRegression(class_weight='balanced', max_iter=1000, n_jobs=-1)"
]
},
"execution_count": 55,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lr.fit(x_train,y_train)"
]
},
{
"cell_type": "code",
"execution_count": 56,
"id": "485d7525-1540-44ac-92e1-b30681920196",
"metadata": {},
"outputs": [],
"source": [
"lrtahmin=lr.predict(x_val)"
]
},
{
"cell_type": "code",
"execution_count": 57,
"id": "b2e4a6c5-0c71-4768-bb52-f2258fa9004d",
"metadata": {},
"outputs": [],
"source": [
"y_val_proba=lr.predict_proba(x_val)[:, 1]"
]
},
{
"cell_type": "code",
"execution_count": 58,
"id": "5c64fce1-0287-4c42-b8ac-6b2821929a97",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"ROC AUC: 0.6949\n"
]
}
],
"source": [
"roc_auc=roc_auc_score(y_val, y_val_proba)\n",
"print(f\"ROC AUC: {roc_auc:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 59,
"id": "442591df-0988-4728-8397-d40c550e862a",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.6265\n"
]
}
],
"source": [
"acc=accuracy_score(y_val, lrtahmin)\n",
"print(f\"Accuracy: {acc:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 60,
"id": "d4db66aa-6fd2-4588-b6a9-285c159c3fea",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([[35940, 16799],\n",
" [35487, 51774]])"
]
},
"execution_count": 60,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"confusion_matrix(y_val, lrtahmin)"
]
},
{
"cell_type": "code",
"execution_count": 61,
"id": "1d008dca-d808-43c8-88eb-cd420ef85097",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Classification Report:\n",
" precision recall f1-score support\n",
"\n",
" 0.0 0.50 0.68 0.58 52739\n",
" 1.0 0.76 0.59 0.66 87261\n",
"\n",
" accuracy 0.63 140000\n",
" macro avg 0.63 0.64 0.62 140000\n",
"weighted avg 0.66 0.63 0.63 140000\n",
"\n"
]
}
],
"source": [
"print(\"\\nClassification Report:\")\n",
"print(classification_report(y_val, lrtahmin))"
]
},
{
"cell_type": "code",
"execution_count": 64,
"id": "31bfb2d9-d9a2-481c-a3ca-76fea0520b7a",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"fpr, tpr, thresholds=roc_curve(y_val, y_val_proba)\n",
"roc_auc=auc(fpr, tpr)\n",
"\n",
"plt.figure(figsize=(6,5))\n",
"plt.plot(fpr, tpr, label=f'Logistic Regression (AUC = {roc_auc:.3f})')\n",
"plt.plot([0, 1], [0, 1], 'k--', label='Random Classifier')\n",
"\n",
"plt.xlabel('False Positive Rate')\n",
"plt.ylabel('True Positive Rate')\n",
"plt.title('ROC Curve')\n",
"plt.legend(loc='lower right')\n",
"plt.grid(alpha=0.3)\n",
"plt.show()"
]
},
{
"cell_type": "code",
"execution_count": 67,
"id": "d7301628-8e45-4b82-9fe2-20be28972956",
"metadata": {},
"outputs": [],
"source": [
"from sklearn.linear_model import LogisticRegression\n",
"from sklearn.neighbors import KNeighborsClassifier\n",
"from sklearn.tree import DecisionTreeClassifier\n",
"from sklearn.ensemble import RandomForestClassifier\n",
"from sklearn.ensemble import AdaBoostClassifier\n",
"from sklearn.ensemble import GradientBoostingClassifier\n",
"from sklearn.naive_bayes import MultinomialNB\n",
"from sklearn.naive_bayes import BernoulliNB\n",
"\n",
"from sklearn.metrics import accuracy_score, precision_score, recall_score\n",
"from sklearn.metrics import f1_score, confusion_matrix, classification_report\n",
"from sklearn.model_selection import train_test_split\n",
"\n",
"b = BernoulliNB()\n",
"l = LogisticRegression()\n",
"d = DecisionTreeClassifier()\n",
"r = RandomForestClassifier()\n",
"gb= GradientBoostingClassifier()\n",
"kn= KNeighborsClassifier()\n",
"ab= AdaBoostClassifier()\n",
"\n",
"def algo_test(x, y):\n",
" modeller=[ b, l, d, r, gb, kn, ab]\n",
" isimler=[\"BernoulliNB\", \"LogisticRegression\", \"DecisionTreeClassifier\", \n",
" \"RandomForestClassifier\", \"GradientBoostingClassifier\", \"KNeighborsClassifier\",\n",
" \"AdaBoostClassifier\"]\n",
"\n",
" x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=.3, random_state = 42)\n",
" \n",
" accuracy = []\n",
" precision = []\n",
" recall = []\n",
" f1 = []\n",
" mdl=[]\n",
"\n",
" print(\"Veriler hazır modeller deneniyor\")\n",
" for model in modeller:\n",
" print(model, \" modeli eğitiliyor!..\")\n",
" model=model.fit(x_train,y_train)\n",
" tahmin=model.predict(x_test)\n",
" mdl.append(model)\n",
" accuracy.append(accuracy_score(y_test, tahmin))\n",
" precision.append(precision_score(y_test, tahmin, average=\"micro\"))\n",
" recall.append(recall_score(y_test, tahmin, average=\"micro\"))\n",
" f1.append(f1_score(y_test, tahmin, average=\"micro\"))\n",
" print(confusion_matrix(y_test, tahmin))\n",
"\n",
" print(\"Eğitim tamamlandı.\")\n",
" \n",
" metrics=pd.DataFrame(columns=[\"Accuracy\", \"Precision\", \"Recall\", \"F1\", \"Model\"], index=isimler)\n",
" metrics[\"Accuracy\"] = accuracy\n",
" metrics[\"Precision\"] = precision \n",
" metrics[\"Recall\"] = recall\n",
" metrics[\"F1\"] = f1\n",
" metrics[\"Model\"]=mdl\n",
"\n",
" metrics.sort_values(\"F1\", ascending=False, inplace=True)\n",
"\n",
" print(\"En başarılı model: \", metrics.iloc[0].name)\n",
" model=metrics.iloc[0,-1]\n",
" tahmin=model.predict(np.array(x_test) if model==kn else x_test)\n",
" print(\"Confusion Matrix:\")\n",
" print(confusion_matrix(y_test, tahmin))\n",
" print(\"classification Report:\")\n",
" print(classification_report(y_test, tahmin))\n",
" print(\"Diğer Modeller:\")\n",
" \n",
" return metrics.drop(\"Model\", axis=1)"
]
},
{
"cell_type": "code",
"execution_count": 68,
"id": "160cd1ff-5537-4344-9863-f3c3fe9b760b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Veriler hazır modeller deneniyor\n",
"BernoulliNB() modeli eğitiliyor!..\n",
"[[ 7905 70919]\n",
" [ 4299 126877]]\n",
"LogisticRegression() modeli eğitiliyor!..\n",
"[[ 23026 55798]\n",
" [ 16367 114809]]\n",
"DecisionTreeClassifier() modeli eğitiliyor!..\n",
"[[36517 42307]\n",
" [45121 86055]]\n",
"RandomForestClassifier() modeli eğitiliyor!..\n",
"[[ 29277 49547]\n",
" [ 21300 109876]]\n",
"GradientBoostingClassifier() modeli eğitiliyor!..\n",
"[[ 27469 51355]\n",
" [ 18027 113149]]\n",
"KNeighborsClassifier() modeli eğitiliyor!..\n",
"[[29112 49712]\n",
" [33933 97243]]\n",
"AdaBoostClassifier() modeli eğitiliyor!..\n",
"[[ 26408 52416]\n",
" [ 18590 112586]]\n",
"Eğitim tamamlandı.\n",
"En başarılı model: GradientBoostingClassifier\n",
"Confusion Matrix:\n",
"[[ 27469 51355]\n",
" [ 18027 113149]]\n",
"classification Report:\n",
" precision recall f1-score support\n",
"\n",
" 0.0 0.60 0.35 0.44 78824\n",
" 1.0 0.69 0.86 0.77 131176\n",
"\n",
" accuracy 0.67 210000\n",
" macro avg 0.65 0.61 0.60 210000\n",
"weighted avg 0.66 0.67 0.64 210000\n",
"\n",
"Diğer Modeller:\n"
]
},
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" Accuracy \n",
" Precision \n",
" Recall \n",
" F1 \n",
" \n",
" \n",
" \n",
" \n",
" GradientBoostingClassifier \n",
" 0.669610 \n",
" 0.669610 \n",
" 0.669610 \n",
" 0.669610 \n",
" \n",
" \n",
" RandomForestClassifier \n",
" 0.662633 \n",
" 0.662633 \n",
" 0.662633 \n",
" 0.662633 \n",
" \n",
" \n",
" AdaBoostClassifier \n",
" 0.661876 \n",
" 0.661876 \n",
" 0.661876 \n",
" 0.661876 \n",
" \n",
" \n",
" LogisticRegression \n",
" 0.656357 \n",
" 0.656357 \n",
" 0.656357 \n",
" 0.656357 \n",
" \n",
" \n",
" BernoulliNB \n",
" 0.641819 \n",
" 0.641819 \n",
" 0.641819 \n",
" 0.641819 \n",
" \n",
" \n",
" KNeighborsClassifier \n",
" 0.601690 \n",
" 0.601690 \n",
" 0.601690 \n",
" 0.601690 \n",
" \n",
" \n",
" DecisionTreeClassifier \n",
" 0.583676 \n",
" 0.583676 \n",
" 0.583676 \n",
" 0.583676 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" Accuracy Precision Recall F1\n",
"GradientBoostingClassifier 0.669610 0.669610 0.669610 0.669610\n",
"RandomForestClassifier 0.662633 0.662633 0.662633 0.662633\n",
"AdaBoostClassifier 0.661876 0.661876 0.661876 0.661876\n",
"LogisticRegression 0.656357 0.656357 0.656357 0.656357\n",
"BernoulliNB 0.641819 0.641819 0.641819 0.641819\n",
"KNeighborsClassifier 0.601690 0.601690 0.601690 0.601690\n",
"DecisionTreeClassifier 0.583676 0.583676 0.583676 0.583676"
]
},
"execution_count": 68,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"algo_test(x,y)"
]
},
{
"cell_type": "markdown",
"id": "2cc84f84-30fa-4ef9-a0ee-53dfeda383e1",
"metadata": {},
"source": [
"##### XGBClassifier"
]
},
{
"cell_type": "code",
"execution_count": 70,
"id": "485d44d5-02d4-4553-92e0-4abcd8cf9f8d",
"metadata": {},
"outputs": [],
"source": [
"xg=XGBClassifier(\n",
" objective='binary:logistic',\n",
" eval_metric='auc',\n",
" n_estimators=1000,\n",
" learning_rate=0.03,\n",
" max_depth=4,\n",
" min_child_weight=5,\n",
" subsample=0.8,\n",
" colsample_bytree=0.8,\n",
" gamma=0,\n",
" reg_alpha=0,\n",
" reg_lambda=1,\n",
" random_state=42,\n",
" n_jobs=-1\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 71,
"id": "d3956c1c-cc09-41fb-b76c-7eb7553d1807",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
" colsample_bylevel=None, colsample_bynode=None,\n",
" colsample_bytree=0.8, device=None, early_stopping_rounds=None,\n",
" enable_categorical=False, eval_metric='auc', feature_types=None,\n",
" feature_weights=None, gamma=0, grow_policy=None,\n",
" importance_type=None, interaction_constraints=None,\n",
" learning_rate=0.03, max_bin=None, max_cat_threshold=None,\n",
" max_cat_to_onehot=None, max_delta_step=None, max_depth=4,\n",
" max_leaves=None, min_child_weight=5, missing=nan,\n",
" monotone_constraints=None, multi_strategy=None, n_estimators=1000,\n",
" n_jobs=-1, num_parallel_tree=None, ...) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. \n",
"
\n",
"
\n",
" Parameters \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" objective \n",
" 'binary:logistic' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" base_score \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" booster \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" callbacks \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" colsample_bylevel \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" colsample_bynode \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" colsample_bytree \n",
" 0.8 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" device \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" early_stopping_rounds \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" enable_categorical \n",
" False \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" eval_metric \n",
" 'auc' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" feature_types \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" feature_weights \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" gamma \n",
" 0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" grow_policy \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" importance_type \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" interaction_constraints \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" learning_rate \n",
" 0.03 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_bin \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_cat_threshold \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_cat_to_onehot \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_delta_step \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_depth \n",
" 4 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_leaves \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" min_child_weight \n",
" 5 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" missing \n",
" nan \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" monotone_constraints \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" multi_strategy \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" n_estimators \n",
" 1000 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" n_jobs \n",
" -1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" num_parallel_tree \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" random_state \n",
" 42 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" reg_alpha \n",
" 0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" reg_lambda \n",
" 1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" sampling_method \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" scale_pos_weight \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" subsample \n",
" 0.8 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" tree_method \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" validate_parameters \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" verbosity \n",
" None \n",
" \n",
" \n",
" \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
"XGBClassifier(base_score=None, booster=None, callbacks=None,\n",
" colsample_bylevel=None, colsample_bynode=None,\n",
" colsample_bytree=0.8, device=None, early_stopping_rounds=None,\n",
" enable_categorical=False, eval_metric='auc', feature_types=None,\n",
" feature_weights=None, gamma=0, grow_policy=None,\n",
" importance_type=None, interaction_constraints=None,\n",
" learning_rate=0.03, max_bin=None, max_cat_threshold=None,\n",
" max_cat_to_onehot=None, max_delta_step=None, max_depth=4,\n",
" max_leaves=None, min_child_weight=5, missing=nan,\n",
" monotone_constraints=None, multi_strategy=None, n_estimators=1000,\n",
" n_jobs=-1, num_parallel_tree=None, ...)"
]
},
"execution_count": 71,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"xg.fit(x_train,y_train)"
]
},
{
"cell_type": "code",
"execution_count": 72,
"id": "e83acd48-70e0-4838-af26-3874df1fcfde",
"metadata": {},
"outputs": [],
"source": [
"xgtahmin=xg.predict(x_val)"
]
},
{
"cell_type": "code",
"execution_count": 73,
"id": "058ec30b-ed64-4d46-88e9-51535632362c",
"metadata": {},
"outputs": [],
"source": [
"y_proba_xg=xg.predict_proba(x_val)[:, 1]"
]
},
{
"cell_type": "code",
"execution_count": 74,
"id": "f006aa06-2215-4db6-b7b8-9ded7b313240",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"ROC AUC: 0.7221\n"
]
}
],
"source": [
"roc_auc=roc_auc_score(y_val, y_proba_xg)\n",
"print(f\"ROC AUC: {roc_auc:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 75,
"id": "7c9cd915-766b-45d3-b37f-4ecb242e2cd1",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.6813\n"
]
}
],
"source": [
"acc=accuracy_score(y_val, xgtahmin)\n",
"print(f\"Accuracy: {acc:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 76,
"id": "3eeffc3a-1bcd-40f4-9116-cbaf6960d6be",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([[21342, 31397],\n",
" [13226, 74035]])"
]
},
"execution_count": 76,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"confusion_matrix(y_val, xgtahmin)"
]
},
{
"cell_type": "code",
"execution_count": 77,
"id": "bb4259f2-fcc3-41cb-bc4e-4a61ec29b889",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Classification Report:\n",
" precision recall f1-score support\n",
"\n",
" 0.0 0.62 0.40 0.49 52739\n",
" 1.0 0.70 0.85 0.77 87261\n",
"\n",
" accuracy 0.68 140000\n",
" macro avg 0.66 0.63 0.63 140000\n",
"weighted avg 0.67 0.68 0.66 140000\n",
"\n"
]
}
],
"source": [
"print(\"\\nClassification Report:\")\n",
"print(classification_report(y_val, xgtahmin))"
]
},
{
"cell_type": "code",
"execution_count": 78,
"id": "1d9e7d78-c3e6-4e73-993d-c84effe41578",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"fpr, tpr, thresholds=roc_curve(y_val, y_proba_xg)\n",
"roc_auc=auc(fpr, tpr)\n",
"\n",
"plt.figure(figsize=(6,5))\n",
"plt.plot(fpr, tpr, label=f'Logistic Regression (AUC = {roc_auc:.3f})')\n",
"plt.plot([0, 1], [0, 1], 'k--', label='Random Classifier')\n",
"\n",
"plt.xlabel('False Positive Rate')\n",
"plt.ylabel('True Positive Rate')\n",
"plt.title('ROC Curve')\n",
"plt.legend(loc='lower right')\n",
"plt.grid(alpha=0.3)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"id": "66fed059-6534-4a3c-86e4-dc0881b300d3",
"metadata": {},
"source": [
"##### LGBMClassifier"
]
},
{
"cell_type": "code",
"execution_count": 99,
"id": "f4f0282e-5647-4722-aa20-dc50344bf6ca",
"metadata": {},
"outputs": [],
"source": [
"lg=LGBMClassifier(\n",
" objective='binary',\n",
" boosting_type='gbdt',\n",
" n_estimators=1000,\n",
" learning_rate=0.03,\n",
" max_depth=-1,\n",
" num_leaves=31,\n",
" subsample=0.8,\n",
" colsample_bytree=0.8,\n",
" reg_alpha=0.0,\n",
" reg_lambda=1.0,\n",
" min_child_samples=20,\n",
" random_state=42,\n",
" n_jobs=-1\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 100,
"id": "e008c82a-b33f-4888-913f-e5bbf0b23372",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[LightGBM] [Warning] Found whitespace in feature_names, replace with underlines\n",
"[LightGBM] [Info] Number of positive: 349046, number of negative: 210954\n",
"[LightGBM] [Info] Auto-choosing row-wise multi-threading, the overhead of testing was 0.014475 seconds.\n",
"You can set `force_row_wise=true` to remove the overhead.\n",
"And if memory is not enough, you can set `force_col_wise=true`.\n",
"[LightGBM] [Info] Total Bins 2501\n",
"[LightGBM] [Info] Number of data points in the train set: 560000, number of used features: 45\n",
"[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.623296 -> initscore=0.503564\n",
"[LightGBM] [Info] Start training from score 0.503564\n"
]
},
{
"data": {
"text/html": [
"LGBMClassifier(colsample_bytree=0.8, learning_rate=0.03, n_estimators=1000,\n",
" n_jobs=-1, objective='binary', random_state=42, reg_lambda=1.0,\n",
" subsample=0.8) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. iFitted
\n",
"
\n",
"
\n",
" Parameters \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" boosting_type \n",
" 'gbdt' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" num_leaves \n",
" 31 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_depth \n",
" -1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" learning_rate \n",
" 0.03 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" n_estimators \n",
" 1000 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" subsample_for_bin \n",
" 200000 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" objective \n",
" 'binary' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" class_weight \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" min_split_gain \n",
" 0.0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" min_child_weight \n",
" 0.001 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" min_child_samples \n",
" 20 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" subsample \n",
" 0.8 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" subsample_freq \n",
" 0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" colsample_bytree \n",
" 0.8 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" reg_alpha \n",
" 0.0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" reg_lambda \n",
" 1.0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" random_state \n",
" 42 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" n_jobs \n",
" -1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" importance_type \n",
" 'split' \n",
" \n",
" \n",
" \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
"LGBMClassifier(colsample_bytree=0.8, learning_rate=0.03, n_estimators=1000,\n",
" n_jobs=-1, objective='binary', random_state=42, reg_lambda=1.0,\n",
" subsample=0.8)"
]
},
"execution_count": 100,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"lg.fit(x_train,y_train)"
]
},
{
"cell_type": "code",
"execution_count": 101,
"id": "9f74ffed-e47a-49dc-be5d-4b5ecddf0a46",
"metadata": {},
"outputs": [],
"source": [
"lgtahmin=lg.predict(x_val)"
]
},
{
"cell_type": "code",
"execution_count": 102,
"id": "7a6064ab-f87d-496b-820b-71e1e0d53300",
"metadata": {},
"outputs": [],
"source": [
"y_proba_lg=lg.predict_proba(x_val)[:, 1]"
]
},
{
"cell_type": "code",
"execution_count": 103,
"id": "b7089902-6b2c-4eaa-88d5-d8a622794e03",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"ROC AUC: 0.7254\n"
]
}
],
"source": [
"roc_auc=roc_auc_score(y_val, y_proba_lg)\n",
"print(f\"ROC AUC: {roc_auc:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 104,
"id": "169bec86-8259-42c5-981a-11adbb25dd2b",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.6839\n"
]
}
],
"source": [
"acc=accuracy_score(y_val, lgtahmin)\n",
"print(f\"Accuracy: {acc:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 105,
"id": "4bf7abba-5763-49d2-99a0-f14460185b1c",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([[22252, 30487],\n",
" [13768, 73493]])"
]
},
"execution_count": 105,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"confusion_matrix(y_val, lgtahmin)"
]
},
{
"cell_type": "code",
"execution_count": 128,
"id": "149c8052-d95f-4d94-a536-6c7141ec256e",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"cm=confusion_matrix(y_val, lgtahmin).astype(int)\n",
"\n",
"sns.heatmap(cm, annot=True, fmt='d')\n",
"plt.show();"
]
},
{
"cell_type": "code",
"execution_count": 88,
"id": "31a17951-e3ce-48dc-a7dc-f36ca3b2fe8c",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"Classification Report:\n",
" precision recall f1-score support\n",
"\n",
" 0.0 0.62 0.42 0.50 52739\n",
" 1.0 0.71 0.84 0.77 87261\n",
"\n",
" accuracy 0.68 140000\n",
" macro avg 0.66 0.63 0.63 140000\n",
"weighted avg 0.67 0.68 0.67 140000\n",
"\n"
]
}
],
"source": [
"print(\"\\nClassification Report:\")\n",
"print(classification_report(y_val, lgtahmin))"
]
},
{
"cell_type": "code",
"execution_count": 89,
"id": "a222f450-46a7-4dae-9514-73436e877989",
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"fpr, tpr, thresholds=roc_curve(y_val, y_proba_lg)\n",
"roc_auc=auc(fpr, tpr)\n",
"\n",
"plt.figure(figsize=(6,5))\n",
"plt.plot(fpr, tpr, label=f'Logistic Regression (AUC = {roc_auc:.3f})')\n",
"plt.plot([0, 1], [0, 1], 'k--', label='Random Classifier')\n",
"\n",
"plt.xlabel('False Positive Rate')\n",
"plt.ylabel('True Positive Rate')\n",
"plt.title('ROC Curve')\n",
"plt.legend(loc='lower right')\n",
"plt.grid(alpha=0.3)\n",
"plt.show()"
]
},
{
"cell_type": "markdown",
"id": "7b1b9126-fb94-4398-9162-de80f0c06dad",
"metadata": {},
"source": [
"### Test"
]
},
{
"cell_type": "code",
"execution_count": 90,
"id": "2e1f9237-8fe9-4b80-8b2c-ab4f2ea69b14",
"metadata": {},
"outputs": [],
"source": [
"sub=pd.read_csv('sample_submission.csv')"
]
},
{
"cell_type": "code",
"execution_count": 91,
"id": "c0b72ba2-5c94-4a5e-9102-a7fbe2eb916a",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" diagnosed_diabetes \n",
" \n",
" \n",
" \n",
" \n",
" 0 \n",
" 700000 \n",
" 0 \n",
" \n",
" \n",
" 1 \n",
" 700001 \n",
" 0 \n",
" \n",
" \n",
" 2 \n",
" 700002 \n",
" 0 \n",
" \n",
" \n",
" 3 \n",
" 700003 \n",
" 0 \n",
" \n",
" \n",
" 4 \n",
" 700004 \n",
" 0 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id diagnosed_diabetes\n",
"0 700000 0\n",
"1 700001 0\n",
"2 700002 0\n",
"3 700003 0\n",
"4 700004 0"
]
},
"execution_count": 91,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"sub.head()"
]
},
{
"cell_type": "code",
"execution_count": 114,
"id": "c35f7caf-cb2a-4144-9780-b784cc52aa92",
"metadata": {},
"outputs": [],
"source": [
"x=train.drop(['diagnosed_diabetes', 'id'], axis=1)"
]
},
{
"cell_type": "code",
"execution_count": 115,
"id": "7b393310-55f7-4096-847a-e668ce04654f",
"metadata": {},
"outputs": [],
"source": [
"y=train['diagnosed_diabetes']"
]
},
{
"cell_type": "code",
"execution_count": 116,
"id": "ec519f5b-b1a6-4c03-94e9-8a9efa6b1ed5",
"metadata": {},
"outputs": [],
"source": [
"x_test=test.drop(['diagnosed_diabetes','id'],axis=1)"
]
},
{
"cell_type": "code",
"execution_count": 117,
"id": "e0db3206-18c8-4888-938e-1184389873d6",
"metadata": {},
"outputs": [],
"source": [
"final_lgbm=LGBMClassifier(\n",
" objective='binary',\n",
" boosting_type='gbdt',\n",
" n_estimators=1000,\n",
" learning_rate=0.03,\n",
" max_depth=-1,\n",
" num_leaves=31,\n",
" subsample=0.8,\n",
" colsample_bytree=0.8,\n",
" reg_alpha=0.0,\n",
" reg_lambda=1.0,\n",
" min_child_samples=20,\n",
" random_state=42,\n",
" n_jobs=-1\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 118,
"id": "b8c86e82-e43b-497d-8fb9-d66452999f60",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[LightGBM] [Warning] Found whitespace in feature_names, replace with underlines\n",
"[LightGBM] [Info] Number of positive: 436307, number of negative: 263693\n",
"[LightGBM] [Info] Auto-choosing row-wise multi-threading, the overhead of testing was 0.030948 seconds.\n",
"You can set `force_row_wise=true` to remove the overhead.\n",
"And if memory is not enough, you can set `force_col_wise=true`.\n",
"[LightGBM] [Info] Total Bins 2497\n",
"[LightGBM] [Info] Number of data points in the train set: 700000, number of used features: 45\n",
"[LightGBM] [Info] [binary:BoostFromScore]: pavg=0.623296 -> initscore=0.503561\n",
"[LightGBM] [Info] Start training from score 0.503561\n"
]
},
{
"data": {
"text/html": [
"LGBMClassifier(colsample_bytree=0.8, learning_rate=0.03, n_estimators=1000,\n",
" n_jobs=-1, objective='binary', random_state=42, reg_lambda=1.0,\n",
" subsample=0.8) In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org. iFitted
\n",
"
\n",
"
\n",
" Parameters \n",
" \n",
" \n",
" \n",
" \n",
" \n",
" boosting_type \n",
" 'gbdt' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" num_leaves \n",
" 31 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" max_depth \n",
" -1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" learning_rate \n",
" 0.03 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" n_estimators \n",
" 1000 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" subsample_for_bin \n",
" 200000 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" objective \n",
" 'binary' \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" class_weight \n",
" None \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" min_split_gain \n",
" 0.0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" min_child_weight \n",
" 0.001 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" min_child_samples \n",
" 20 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" subsample \n",
" 0.8 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" subsample_freq \n",
" 0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" colsample_bytree \n",
" 0.8 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" reg_alpha \n",
" 0.0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" reg_lambda \n",
" 1.0 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" random_state \n",
" 42 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" n_jobs \n",
" -1 \n",
" \n",
" \n",
"\n",
" \n",
" \n",
" importance_type \n",
" 'split' \n",
" \n",
" \n",
" \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
"LGBMClassifier(colsample_bytree=0.8, learning_rate=0.03, n_estimators=1000,\n",
" n_jobs=-1, objective='binary', random_state=42, reg_lambda=1.0,\n",
" subsample=0.8)"
]
},
"execution_count": 118,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"final_lgbm.fit(x, y)"
]
},
{
"cell_type": "code",
"execution_count": 120,
"id": "cd2ea801-7be2-4239-ae5a-da1171709fe5",
"metadata": {},
"outputs": [],
"source": [
"test_proba=final_lgbm.predict_proba(x_test)[:, 1]"
]
},
{
"cell_type": "code",
"execution_count": 121,
"id": "81ebb5c5-717d-4933-8e20-041b4dcaec54",
"metadata": {},
"outputs": [],
"source": [
"submission=pd.DataFrame({'id':df2['id'],'diagnosed_diabetes':test_proba})"
]
},
{
"cell_type": "code",
"execution_count": 122,
"id": "4e16ee91-1b9a-4799-a393-01090076841c",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" \n",
" id \n",
" diagnosed_diabetes \n",
" \n",
" \n",
" \n",
" \n",
" 0 \n",
" 700000 \n",
" 0.507012 \n",
" \n",
" \n",
" 1 \n",
" 700001 \n",
" 0.685015 \n",
" \n",
" \n",
" 2 \n",
" 700002 \n",
" 0.768543 \n",
" \n",
" \n",
" 3 \n",
" 700003 \n",
" 0.398964 \n",
" \n",
" \n",
" 4 \n",
" 700004 \n",
" 0.918060 \n",
" \n",
" \n",
"
\n",
"
"
],
"text/plain": [
" id diagnosed_diabetes\n",
"0 700000 0.507012\n",
"1 700001 0.685015\n",
"2 700002 0.768543\n",
"3 700003 0.398964\n",
"4 700004 0.918060"
]
},
"execution_count": 122,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"submission.head()"
]
},
{
"cell_type": "code",
"execution_count": 123,
"id": "29df8395-6c34-4192-b1c1-74536f171e94",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"np.True_"
]
},
"execution_count": 123,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"submission['diagnosed_diabetes'].between(0, 1).all()"
]
},
{
"cell_type": "code",
"execution_count": 124,
"id": "205d471e-0910-4705-9a24-e71d2c60c959",
"metadata": {},
"outputs": [],
"source": [
"submission.to_csv('submission.csv',index=False)"
]
},
{
"cell_type": "code",
"execution_count": 125,
"id": "9fcb200a-579a-45dd-9e67-ba67de997003",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"(300000, 2)\n"
]
}
],
"source": [
"print(submission.shape)"
]
},
{
"cell_type": "code",
"execution_count": 126,
"id": "6132674d-de4f-43d3-983c-ae8344f3c59a",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.059118226321828095 0.985593126754481\n"
]
}
],
"source": [
"print(submission['diagnosed_diabetes'].min(), submission['diagnosed_diabetes'].max())"
]
},
{
"cell_type": "code",
"execution_count": 127,
"id": "51a27953-9584-4451-a994-91698f1b82d6",
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"id 0\n",
"diagnosed_diabetes 0\n",
"dtype: int64\n"
]
}
],
"source": [
"print(submission.isna().sum())"
]
},
{
"cell_type": "code",
"execution_count": 129,
"id": "1b6fce58-917c-43b7-9d8e-290a7cdebe95",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['lr_final_model.pkl']"
]
},
"execution_count": 129,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"joblib.dump(final_lgbm,'lr_final_model.pkl')"
]
},
{
"cell_type": "code",
"execution_count": 130,
"id": "f4e15d28-b62d-43f1-984d-084f445e4456",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"['feature_names.pkl']"
]
},
"execution_count": 130,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"joblib.dump(x_train.columns.tolist(),'feature_names.pkl')"
]
},
{
"cell_type": "markdown",
"id": "a0186124-209b-4770-a39f-83b3a5d42147",
"metadata": {},
"source": [
"### Conclusion\n",
"The final LightGBM model achieved a ROC-AUC of approximately 0.73, indicating a solid ability to distinguish between patients with and without diabetes."
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python [conda env:base] *",
"language": "python",
"name": "conda-base-py"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.13.5"
}
},
"nbformat": 4,
"nbformat_minor": 5
}