This repository contains the solution for our group in the competition of predicting the price of a bottle of wine based on a collection of over one hundred thousand reviews and other product features.
The data description is given in kaggle like below:
train.csv - the training set
test.csv - the test set
country (String): The country that the wine is from.
province (String): The province or state that the wine is from.
region_1 (String): The wine growing area in a province or state.
region_2 (String): Sometimes there are more specific regions within the wine growing area.
winery (String): The winery that made the wine.
variety (String): The type of grapes used to make the wine.]
designation (String): The vineyard within the winery where the grapes that made the wine are from.
taster_name (String): taster name.
taster_twitter_handle (String): taster twitter account name.
review (String): A few sentences from a sommelier describing the wine's taste, smell, look, feel, etc.
review_score (Numeric): Number of points WineEnthusiast rated the wine on a scale of 1-100.
TARGET: price (Numeric) The cost for a bottle of wine.