Machine Learning Board Game Analysis and Prediction

This repository contains Machine Learning and Exploratory Data Analysis (EDA) of a dataset from Ludii board games Ludii.games general game system, focusing on predicting board game attributes. The project includes regression and classification tasks, along with preprocessing, feature engineering, and evaluation.

Objectives

• Predict numerical attributes like OriginYear and UCT.
• Classify categorical labels like Category, Region, and BestAgent.

Key Techniques

• PCA for dimensionality reduction
• Outlier detection using K-Means clustering
• Model training with Random Forest and Stacking regressors/classifiers
• Hyperparameter tuning and cross-validation

Prerequisites

• Python 3.x
• Pip: Python package installer, or alternatively Conda.
• Libraries:
  • pandas
  • numpy
  • scikit-learn
  • matplotlib
  • seaborn
  • Jupyter Notebook
• GameData.csv: Sourced from Ludii.games. Clone the repository to access the dataset.

You can install the required libraries using pip.

License

This project is licensed under the MIT License.