My Graduate Capstone Project - This is a Product Recommendation System for a Local Wholesaler in India, using Python and Machine Learning
Dataset :- About two and half years of data used for this project and is uploaded on GitHub (Sales Transactions-2017.csv, Sales Transactions-2018.csv, Sales Transactions-2019.csv)
Used Flask web framework in Python to publish the results on a web page
Link to the web application :- https://product-recommendation-system.herokuapp.com/
About the Project :- https://youtu.be/0FCxHEc_e8Q (This video provides an understanding about the company and the project)
PGM-1 - Data_Cleaning.ipynb - Performs Data Cleaning to remove the irrelevant data
PGM-2 - Product_Ranking.ipynb - Identifies the Top Selling and Most Popular products
PGM-3 - Customer-Product-Ranking.ipynb - Identifies the products of a Customer, Most frequently purchased and Purchased the Most
PGM-4 - Recommend_Products_to_Customer.ipynb - Identifies the similarities between the customers. This is further used to recommend the products to a Customer, based on the products they purchased and similarities with other Customer purchase pattern
PGM-5 - Recommend_Similar_Products.ipynb - Identifies the similarities between the products. This correlation is further used to identify the similar products for the item being viewed
These programs create .pkl files (prod_ranking_model.pkl, cust_prod_ranking_model.pkl, cust_correlation_model.pkl, prod_correlation_model.pkl), which are further accessed for publishing data on website
Flask.ipynb - Python program with different functionalities based on the user action on website, to publish the data on to website