app.py

import streamlit as st
import pickle
import numpy as np
import joblib
# Load the trained model
#model = pickle.load(open('credit_score.pkl', 'rb'))
model = joblib.load('credit_score.joblib')

# Define the Streamlit app
st.title("Credit Score Prediction")

# Get user input
annual_income = st.number_input("Annual Income", min_value=0,max_value=100000, step=100)
monthly_salary = st.number_input("Monthly In-hand Salary", min_value=0,max_value=5000,  step=100.0)
num_bank_accounts = st.number_input("Number of Bank Accounts", min_value=0, step=1)
num_credit_cards = st.number_input("Number of Credit Cards", min_value=0, step=1)
interest_rate = st.number_input("Interest Rate (%)", min_value=0.0, max_value=30.0, step=1)
num_loans = st.number_input("Number of Loans", min_value=0, max_value=10,step=1)
avg_days_delayed = st.number_input("Average Number of Days Delayed", min_value=0.0, step=1.0)
num_delayed_payments = st.number_input("Number of Delayed Payments", min_value=0, step=1)
credit_mix = st.selectbox("Credit Mix", ["Bad", "Standard", "Good"])
outstanding_debt = st.number_input("Outstanding Debt", min_value=0.0, step=100.0)
credit_history_age = st.number_input("Credit History Age (Years)", min_value=0.0, step=1.0)
monthly_balance = st.number_input("Monthly Balance", min_value=0.0, step=100.0)

# Convert user input to a numpy array
features = np.array([[annual_income, monthly_salary, num_bank_accounts, num_credit_cards,
                     interest_rate, num_loans, avg_days_delayed, num_delayed_payments,
                     credit_mix.index(), outstanding_debt, credit_history_age, monthly_balance]])

# Make the prediction
class_names = ['Bad', 'Standard', 'Good']
prediction = class_names[int(model.predict(features)[0])]

# Display the prediction
if prediction == 'Good':
    st.success(f"Predicted Credit Score: {prediction} 😊")
elif prediction == 'Standard':
    st.warning(f"Predicted Credit Score: {prediction} 🤔")
else:
    st.error(f"Predicted Credit Score: {prediction} 😞")