app.py

import streamlit as st
import pandas as pd
import joblib

# Model'i yükle
@st.cache_resource
def load_model():
    return joblib.load('click_predict_logistic_regression.joblib')

model = load_model()

st.title('Ad Click Prediction')

st.write('Adjust the sliders below to predict if a user will click on an ad:')

# Kullanıcı girişleri
daily_time = st.slider('Daily Time Spent on Site (minutes)', min_value=32.6, max_value=91.43, value=65.0, step=0.1)
age = st.slider('Age', min_value=19, max_value=61, value=36, step=1)
area_income = st.slider('Area Income ($)', min_value=13996, max_value=79485, value=55000, step=100)
daily_internet = st.slider('Daily Internet Usage (minutes)', min_value=104.78, max_value=269.96, value=180.0, step=0.1)
gender = st.radio('Gender', ['Male', 'Not Male'])

if st.button('Predict'):
    # Girişleri bir DataFrame'e dönüştür
    input_data = pd.DataFrame({
        'Daily Time Spent on Site': [daily_time],
        'Age': [age],
        'Area Income': [area_income],
        'Daily Internet Usage': [daily_internet],
        'Male_not_male': [1 if gender == 'Not Male' else 0]
    })

    # Tahmin yap
    prediction = model.predict(input_data)
    probability = model.predict_proba(input_data)[0][1]

    if prediction[0] == 1:
        st.success(f'This user is likely to click on the ad. Probability: {probability:.2f}')
    else:
        st.error(f'This user is unlikely to click on the ad. Probability: {probability:.2f}')

    # Girdi değerlerini göster
    st.write('Input values:')
    st.write(input_data)