Problem statement:
How can we use the knowledge of the crowd and apply machine learning algorithms to recommend the best restaurants to the customers of the food application. We are creating a restaurant recommender, which can give a user an option that now stands out, so that the user will not receive the same recommendation in case a previous choice was unappealing.
The algorithm being used is collaborative filtering.
Collaborative filtering is of 2 types:
1.) Item-Item Collaborative Filtering
2.) User-User Collaborative Filtering
To evaluate our model we are using K-Fold Validation Tedchnique.