Jimut123 / FML_Project

CS-725 Foundations of Machine Learning Project, IIT Bombay

---

Topic: Solving the Cold Start problem in Recommendation Systems - Case Study on MovieLens Dataset

Team Members: Prateek Chanda (22D0362), Sandarbh Yadav (22D0374), Jimut Bahan Pal (22D1594) and Goda Nagakalyani (214050010)

Instructor: Preethi Jyothi

Demo using Gradio:

Gradio demonstration links: MovieLens Dataset Analysis & MovieLens Final Top K recommendations

Instructions for running the code:

1. Install tensorflow, numpy, pandas etc.
2. Clone the repository
3. Run mainSol.py using command: python mainSol.py

To run lightGCN++, in mainSol, pass argument defaultMode=False, else to run LightGCN, pass argument defaultMode=True

Note: LightGCN.py contains the implementation of LightGCN model and recommendation.py contains the gradio code for making recommendations.

Code for HuggingSpace Deployments

Top K recommendation code HuggingSpace

Movielens Dataset analysis code HuggingSpace

---