pages/THE ENSEMBLE METHODS.py

import streamlit as st

st.set_page_config(page_title="Bagging & Random Forest")

st.title("🌟 Ensemble Learning: Bagging & Random Forest")

# Topic selector
technique = st.radio("🔍 Choose a Technique", [
    "👜 Bagging (Bootstrap Aggregation)",
    "🌳 Random Forest"
], horizontal=True)

if technique == "👜 Bagging (Bootstrap Aggregation)":
    st.header("👜 Bagging (Bootstrap Aggregation)")
    st.markdown("""
    **Bagging** stands for *Bootstrap Aggregating*. It is an ensemble method that improves model stability and accuracy by training multiple models on different subsets of the data.

    ### 🔍 How it Works:
    - Multiple models are trained on different **bootstrapped samples** (random samples with replacement).
    - Each model gives a prediction, and the results are **aggregated** (e.g., majority vote for classification, average for regression).

    ### ✅ Use Cases:
    - High-variance models (e.g., Decision Trees)
    - When overfitting is a concern

    ### 🔧 Key Parameters:
    - `n_estimators`: Number of base models
    - `max_samples`: Number of samples per model
    - `bootstrap`: Whether to sample with replacement

    ### ✅ Pros:
    - Reduces variance
    - Prevents overfitting
    - Easy to parallelize

    ### ⚠️ Cons:
    - Can be computationally expensive
    - Doesn't reduce bias
    """)

elif technique == "🌳 Random Forest":
    st.header("🌳 Random Forest")
    st.markdown("""
    **Random Forest** is a popular ensemble method built on top of Bagging, with an extra layer of randomness.

    ### 🔍 How it Works:
    - Uses **Bagging** with Decision Trees.
    - Adds randomness by selecting a random subset of features at each split (not just data samples).

    ### ✅ Use Cases:
    - Classification and Regression
    - Feature selection and importance scoring

    ### 🔧 Key Parameters:
    - `n_estimators`: Number of trees in the forest
    - `max_depth`: Maximum tree depth
    - `max_features`: Number of features to consider at each split
    - `bootstrap`: Whether to use bootstrapped samples

    ### ✅ Pros:
    - Reduces variance and overfitting
    - Handles high-dimensional data well
    - Provides feature importance

    ### ⚠️ Cons:
    - Slower for large forests
    - Less interpretable than single decision trees
    """)

# Common footer note
st.markdown("""
---
📌 **Tip**: Bagging works best with models that have high variance, and Random Forest is a powerful extension of this idea that works exceptionally well on structured/tabular data.
""")