Zomato Restaurant Rating Prediction

Data Mining Project - 2019

In this project the Bangalore restaurant data analysis is done. Main motive of this project is to get a model to predict the rating of the restaurdent rating using the features of the restaurant like menu items, location, price range etc.
So, in this dataset is used of Kaggle. For the analysis of the data it's 3 versions are used:

1. Original Dataset.
2. By Dropping least useful attributes.
3. Class Balanced Dataset.

Tasks

So, in this two tasks have been done first where reviewer's data have generated and another where reviewer's data haven't been taken.

Feature Engineering

Following Methods have been used:

1. Voting Feature Selector using LinearRegression, RFE, Ridge, Lasso & RandomForestRegressor and dropping the attributes whose Mean Ranking is <= 0.35.
2. Using Xgboost feature importance ranking.

Results:

All Right Reserved to Repository Owner.