DSC550: Data Mining

Project Overview

Advanced data mining techniques including association rules, clustering, classification, and pattern discovery for extracting insights from large datasets.

Course Information

• Course: DSC550 - Data Mining
• Institution: Bellevue University
• Semester: Fall 2024

Project Structure

project7-dsc550/
├── src/                    # Source code
│   ├── clustering/        # Clustering algorithms
│   ├── association/       # Association rule mining
│   ├── classification/    # Classification models
│   └── pattern_mining/    # Pattern discovery scripts
├── docs/                  # Documentation and reports
│   ├── final_report.pdf   # Final project report
│   ├── mining_results.pdf # Data mining findings
│   └── methodology.md     # Technical methodology
├── input/                 # Input datasets
├── output/                # Mining results and patterns
└── demo/                  # Interactive mining demos

Key Features

• Clustering analysis (K-means, Hierarchical)
• Association rule mining
• Classification algorithms
• Pattern discovery and analysis

Technologies Used

• Python (Scikit-learn, MLxtend)
• Weka data mining toolkit
• R (arules, cluster packages)
• Apriori algorithm implementation
• Visualization libraries

Installation & Setup

# Clone the repository
cd project7-dsc550

# Install dependencies
pip install -r requirements.txt

# Run clustering analysis
python src/clustering_analysis.py

# Mine association rules
python src/association_mining.py

Results

• 28% increase in cross-selling through discovered patterns
• 12 high-confidence association rules identified (support >0.3)
• Customer segmentation with 85% cluster purity
• Optimized inventory management reducing waste by 15%

Author

Komal Shahid - DSC550 Data Mining Project

License

Academic Use Only - Bellevue University