Built 3 different types of recommendation system to predict the ratings/stars for given user ids and business ids. Used the validation dataset to evaluate the accuracy of recommendation systems.
-
Item-based CF recommendation system
Implemented item-based recommendation system using Pearson similarity formula.
-
Model-based recommendation system
Implemented Model-based recommendation system using XGBregressor(a regressor based on the decision tree) to train a model and use the validation dataset to validate your result.
-
Hybrid recommendation system
Implemented Hybrid recommendation system by taking weighted average of item-based CF and Model-based recommendation system by using the below formula:
final score = 𝛼 × score𝑖𝑡𝑒𝑚_𝑏𝑎𝑠𝑒𝑑 + (1−𝛼) × 𝑠𝑐𝑜𝑟𝑒𝑚𝑜𝑑𝑒𝑙_𝑏𝑎𝑠𝑒𝑑
Link to Data files: https://drive.google.com/drive/folders/1kdQlFvqEKkQUXv3JmpH2fCS2NPS_Q5tw
I used RMSE to check the accuracy of my recommendation systems.
Type of Recommendation System RMSE Item-based CF recommendation system: 1.05 Model-based recommendation system: 0.99 Hybrid recommendation system: 0.98