Machine Learning-Powered Netflix Recommendation System

Overview

This project builds a machine learning-based recommendation system for Netflix movies. It explores clustering techniques and similarity measures to recommend movies based on textual and numerical features. Note: code was originally processed in Google Colab, cells may require adjustment for correct execution on local systems.

Features

• Data Preprocessing: Cleans and processes movie metadata.
• Text Analysis: Tokenization, stopword removal, and TF-IDF vectorization.
• Clustering Models: Uses K-Means and Agglomerative Clustering.
• Similarity Metrics: Computes cosine similarity between movie features.
• Visualization: PCA, Silhouette Score, Dendrograms for clustering analysis.

Technologies Used

• Python
• Libraries: numpy, pandas, matplotlib, seaborn, scikit-learn, nltk, statsmodels, yellowbrick

Setup

1. Install dependencies:

   pip install -r requirements.txt

2. Load the dataset (place Netflix.csv in the working directory).
3. Run the Jupyter Notebook:

   jupyter notebook NetflixRecommendationSystem.ipynb

Usage

• Run the notebook to preprocess the dataset and build the clustering models.
• Analyze recommendations based on cosine similarity scores.
• Visualize cluster groupings and assess model performance.