Chess Opening Classifier

📌 Objective

This project aims to build classification models that predict the chess opening based on the first few moves. The goal is to analyze move patterns and use machine learning to make accurate predictions.

📂 Dataset

The dataset was sourced from Kaggle - All Chess Openings.
It consists of recorded chess games with the following structure:

• Target Variable (Opening): The name of the chess opening played.
• Features (Predictors):
  • move1w, move1b – First moves by white and black.
  • move2w, move2b – Second moves by white and black.
  • move3w, move3b – Third moves by white and black.
  • move4w, move4b – Fourth moves by white and black.
  • Colour – Indicates if the player is white or black.

🏆 Models Used

We explore multiple classifiers:

1. Random Forest – Ensemble learning for robust classification.
2. Support Vector Machine (SVM) – Finds optimal decision boundaries.
3. XGBoost – Boosted trees for superior predictive performance.
4. K-Nearest Neighbors (KNN) – Measures similarity to classify openings.
5. Logistic Regression – Simple probabilistic classification.
6. Neural Network (MLPClassifier) – Deep learning-based model.

🚀 Features

• Encodes chess moves using Label Encoding for numerical representation.
• Use multiple classifiers to Understand how each approache Classification.
• Provides visualizations for model evaluation and feature importance.

🏁 Model Visualizations

Each trained classifier is visualized to understand its decision-making process:

Random Forest - Decision Tree Visualization

SVM - Decision Boundary

XGBoost - Feature Importance

KNN - Classification Boundaries

Logistic Regression - Probability Distribution

Neural Network - Training Loss Curve

🔧 Installation & Setup

Click COLAP Link And Downalod Dataset then You are Ready to start