🏀 NBA Draft Combine Data Analysis

Data analysis of NBA Draft Combine statistics using machine learning techniques.

📋 About

Project developed for the Data Exploration (IP2) course at the Master's program, Faculty of Mathematics, University of Belgrade.

The analysis includes:

• Exploratory Data Analysis (EDA)
• Anomaly Detection (Isolation Forest, LOF)
• Dimensionality Reduction (PCA, t-SNE)
• Clustering (K-Means, Hierarchical, DBSCAN)
• Classification (Random Forest, SVM, Neural Networks)
• Association Rules

📊 Key Results

• 2 player archetypes - identified through clustering
• 10% anomalies - 121 players with unusual characteristics
• 98.77% accuracy - Random Forest classifier
• 90% variance - preserved in 10 PCA components

📄 Documentation

Full project documentation in Serbian is available in nba_projekat_final.pdf and includes:

• Data description and methodology
• Detailed analysis of all techniques
• Visualizations and interpretation
• Conclusions and recommendations

🚀 Getting Started

Prerequisites

• Python 3.9+
• Jupyter Notebook

Installation and Running

# Install dependencies
pip install pandas numpy scikit-learn matplotlib seaborn scipy mlxtend jupyter

# Run the notebook
jupyter notebook nba_data_analysis.ipynb

🛠 Technologies

Analysis

• Python
• Jupyter Notebook
• scikit-learn
• pandas, numpy
• matplotlib, seaborn

Documentation

• LaTeX

📈 Dataset

• Source: NBA Database (Kaggle)
• Size: 1,633 players x 61 features

👤 Author

• Anđela Jovanović - @andjixi