すしですか？？: A Collaborative Filtering Approach to Creating Recommendations from Sushi User-Preference Data
I analyzed sushi preferences of people and made suggestions on what they should try next using user-based collaborative filering!
User-based collaborative filtering is a recommendation system that seeks to recommend items to a target user based on recommended items of similar users. This technique was used in the sushi dataset to gather recommendations. The dataset consisted of top 5 sushi ranking preference per user and these were used as a basis of recommendations. The surprise scikit library was used utilizing the KNNBaseLine algorithm which gave the least RMSE of 1.141373. Aside from explicit ratings, a ranking preference can be used as a basis for the creation of user-based collaborative recommender systems.