Market-Basket-Analysis

This project focuses on mining association rules from transactional datasets using three popular algorithms: Apriori, FP-Growth, and CLOSET. The aim is to uncover hidden patterns, frequent itemsets, and meaningful associations between items in large datasets, which are essential techniques in market basket analysis and recommendation systems.

🧠 Algorithms Implemented

• Apriori Algorithm: A classic algorithm that uses a breadth-first search strategy and pruning to efficiently discover frequent itemsets.
• FP-Growth (Frequent Pattern Growth): An advanced algorithm that constructs a compact FP-tree to mine frequent patterns without candidate generation.
• CLOSET (Closed Itemset Mining): A depth-first approach to efficiently mine closed frequent itemsets, which are more compact and informative.

📊 Key Features

• Generation of frequent itemsets using all three algorithms.
• Extraction of strong association rules based on support and confidence thresholds.
• Comparative analysis of the algorithms based on performance and output.
• Clear visualization of itemsets and rules.
• Modular and extensible codebase suitable for integration with larger data mining pipelines.

🛠 Technologies Used

• Python
• mlxtend and custom implementation
• Jupyter Notebook for experimentation and demonstration

🚀 Getting Started

1. Clone the repository:

   git clone https://github.com/shubhro2002/Market-Basket-Analysis.git
   cd Market-Basket-Analysis
   

2. Install required dependencies:

   pip install -r requirements.txt

3. Run the Jupyter notebooks or Python scripts to start mining itemsets and generating rules