File size: 6,247 Bytes
3a794b5 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 | {
"nbformat": 4,
"nbformat_minor": 0,
"metadata": {
"colab": {
"provenance": []
},
"kernelspec": {
"name": "python3",
"display_name": "Python 3"
},
"language_info": {
"name": "python"
}
},
"cells": [
{
"cell_type": "markdown",
"source": [
"Using regular logistic Regression (Highest Accuracy estimate(so far): 83%"
],
"metadata": {
"id": "MzqDlYenTCb_"
}
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"id": "VycaGV_zrYz7",
"collapsed": true
},
"outputs": [],
"source": [
"import tensorflow as tf\n",
"import numpy as np\n",
"import pandas as pd\n",
"from tensorflow import keras\n",
"from keras import layers\n",
"from sklearn.model_selection import train_test_split\n",
"import matplotlib.pyplot as plt\n",
"from tensorflow.keras.utils import to_categorical\n",
"from sklearn.preprocessing import MinMaxScaler\n",
"d = pd.read_csv(\"student_lifestyle_dataset.csv\")\n",
"Y = d[\"Stress_Level\"]\n",
"dic = {\"Moderate\": 1, \"High\": 2, \"Low\": 0}\n",
"for i, element in np.ndenumerate(Y):\n",
" Y[i,] = dic[element]\n",
"X = d.drop(\"Stress_Level\", axis=1)\n",
"scaler = MinMaxScaler()\n",
"X = scaler.fit_transform(X)\n",
"x_train, x_test, y_train, y_test = train_test_split(X, Y, test_size=0.2, train_size=0.8)\n",
"y_train_one_hot = to_categorical(y_train, num_classes=3)\n",
"y_test_one_hot = to_categorical(y_test, num_classes=3)\n",
"w = [np.zeros(x_train.shape[1]) for i in range(3)]\n",
"b = np.array([0.0 for i in range(3)])\n",
"\n",
"def softmax(z):\n",
" e_z = np.exp(z - np.max(z))\n",
" return e_z/sum(e_z)\n",
"\n",
"def forwardPass(weights, x, biases):\n",
" z = np.array([np.dot(w, x) + b for w, b in zip(weights, biases)])\n",
" a = softmax(z)\n",
" return z, a\n",
"\n",
"def cross_entropy(y, y_pred):\n",
" return -np.sum(y * np.log(y_pred + 1e-4))\n",
"\n",
"def backprop(x, y, a):\n",
" dz = a-y\n",
" dw = [dz[i] * x for i in range(len(dz))]\n",
" db = dz\n",
" return dw, db\n",
"\n",
"epochs = 2000\n",
"alpha = 0.01\n",
"for epoch in range(epochs):\n",
" total_loss = 0\n",
" for x, y in zip(x_train, y_train_one_hot):\n",
" z, a = forwardPass(w, x, b)\n",
" loss = cross_entropy(y, a)\n",
" total_loss += loss\n",
" dw, db = backprop(x, y, a)\n",
" for i in range(3):\n",
" w[i] -= alpha * dw[i]\n",
" b[i] -= alpha * db[i]\n",
" print(\"Loss at Epoch\" + str(epoch +1)+ \": \" + str(total_loss/len(x_train)))"
]
},
{
"cell_type": "code",
"source": [
"def predict(x, w, b):\n",
" _, a = forwardPass(w, x, b)\n",
" return np.argmax(a)\n",
"\n",
"predictions = [predict(x, w, b) for x in x_test]\n",
"from sklearn.metrics import accuracy_score\n",
"y_test = np.array(y_test).astype(np.int64)\n",
"predictions = np.array(predictions).astype(np.int64)\n",
"accuracy = accuracy_score(y_test, predictions)\n",
"print(\"Accuracy on test set:\", accuracy)"
],
"metadata": {
"id": "90O27eYWOBeB"
},
"execution_count": null,
"outputs": []
},
{
"cell_type": "markdown",
"source": [
"Using Neural Network Model Highest Accuracy estimate (so far): 98%-97%"
],
"metadata": {
"id": "zbK7PhT-TZXw"
}
},
{
"cell_type": "code",
"source": [
"import tensorflow as tf\n",
"import numpy as np\n",
"import pandas as pd\n",
"from tensorflow import keras\n",
"from keras import layers\n",
"from sklearn.model_selection import train_test_split\n",
"import matplotlib.pyplot as plt\n",
"from tensorflow.keras.utils import to_categorical\n",
"d = pd.read_csv(\"student_lifestyle_dataset.csv\")\n",
"from sklearn.preprocessing import MinMaxScaler\n",
"from tensorflow.keras.optimizers.schedules import ExponentialDecay\n",
"from tensorflow.keras.optimizers import Adam\n",
"Y = d[\"Stress_Level\"]\n",
"X = d.drop(\"Stress_Level\", axis=1)\n",
"scaler = MinMaxScaler()\n",
"X = scaler.fit_transform(X)\n",
"\n",
"dic = {\"Moderate\": 1, \"High\": 2, \"Low\": 0}\n",
"print(Y.shape)\n",
"for i, element in np.ndenumerate(Y):\n",
" Y[i,] = dic[element]\n",
"print(Y)\n",
"x_train, x_test, y_train, y_test = train_test_split(X, Y, test_size=0.2, train_size=0.8)\n",
"\n",
"y_train = y_train.astype(np.int64)\n",
"y_test = y_test.astype(np.int64)\n",
"model = keras.Sequential(\n",
" [\n",
" layers.Dense(activation=\"relu\", units=125, kernel_regularizer=keras.regularizers.l2(0.01)),\n",
" layers.Dense(activation=\"relu\", units=75, kernel_regularizer=keras.regularizers.l2(0.01)),\n",
" layers.Dense(activation=\"relu\", units=15, kernel_regularizer=keras.regularizers.l2(0.01)),\n",
" layers.Dense(activation=\"softmax\", units=3)\n",
" ]\n",
")\n",
"\n",
"\n",
"\n",
"optimizer = Adam(learning_rate=0.001)\n",
"\n",
"model.compile(\n",
" loss=tf.keras.losses.SparseCategoricalCrossentropy(),\n",
" optimizer=optimizer,\n",
" metrics=['accuracy']\n",
")\n",
"\n",
"model.fit(\n",
" x_train,\n",
" y_train,\n",
" validation_data=(x_test, y_test),\n",
" epochs=200\n",
")\n"
],
"metadata": {
"id": "A1mAPsTfO-v3"
},
"execution_count": null,
"outputs": []
}
]
} |