Random-forest-classifier

🌲 Random Forest Classifier Dashboard

Social Network Ads – Machine Learning Project

This project demonstrates a complete end-to-end Machine Learning workflow using a Random Forest Classifier, along with an interactive Streamlit dashboard for real-time prediction and visualization.

The model predicts whether a customer will purchase a product based on:

• Age
• Estimated Salary

---

📌 Project Highlights

• 🔹 Random Forest Classification (Entropy Criterion)
• 🔹 Feature Scaling using StandardScaler
• 🔹 Confusion Matrix Visualization
• 🔹 ROC Curve & AUC Score
• 🔹 Bias–Variance Analysis
• 🔹 Decision Boundary Visualization (Train & Test)
• 🔹 Model Serialization using Pickle
• 🔹 Interactive Streamlit Dashboard with Dark UI

---

📂 Dataset

Dataset: Social Network Ads
Features Used:

• Age
• Estimated Salary

Target Variable:

• Purchased (0 = No, 1 = Yes)

---

🧠 Machine Learning Workflow

1. Data Loading & Preprocessing
2. Train–Test Split (80% / 20%)
3. Feature Scaling
4. Random Forest Model Training
5. Model Evaluation
   • Accuracy
   • Confusion Matrix
   • ROC Curve & AUC
   • Bias vs Variance
6. Decision Boundary Visualization
7. Model & Scaler Serialization
8. Streamlit Dashboard Deployment

---

📊 Model Performance

• Accuracy Score: Displayed in Dashboard
• Bias: Training Score
• Variance: Test Score
• AUC Score: ROC Curve

---

📈 Visualizations Included

• Confusion Matrix
• ROC Curve
• Decision Boundary (Training Set)
• Decision Boundary (Test Set)

---

🖥️ Streamlit Dashboard Features

• 🎚️ Sidebar sliders for Age & Salary
• 🔮 Real-time Purchase Prediction
• 📊 Model Metrics Display
• 📈 ROC Curve Visualization
• 🧮 Confusion Matrix
• 🌙 Modern Dark-Theme UI

---

🛠️ Technologies Used

• Python
• NumPy
• Pandas
• Matplotlib
• Scikit-learn
• Streamlit
• Pickle

---

📌 Future Enhancements

• Add Cross-Validation
• Hyperparameter Tuning
• Multiple Classifiers Comparison
• Deployment on Streamlit Cloud / Hugging Face Spaces
• Feature Importance Visualization