Red Wine Quality

Problem Discussion
#1

Wine is an alcoholic drink typically made from fermented grapes. Yeast consumes the sugar in the grapes and converts it to ethanol, carbon dioxide, and heat.

This is a companion discussion topic for the original entry at https://dphi.tech/practice/challenge/10
1 Like
#2

In Predict the Quality of Red Wine practice problem the quality below 5 is 3.93 percent only whereas above 5 is 53.44 percent. How to handle this imbalance in the dataset? Since here in output variable, the range of quality is from 0 to 10 and our dataset has distrbution
5 : 477
6 : 446
7 : 139
4 : 37
8 : 13
3 : 7

It does not have 0, 1, 2, 9, 10

#3

You can use SMOTE. Also you can take a look to the following: https://www.kaggle.com/questions-and-answers/93669 and https://datascience.stackexchange.com/questions/24610/smote-and-multi-class-oversampling

#4

What accuracy did you receive? I have been receiving extremely poor accuracy using SMOTE.

#5

I used xgboost.XGBRFClassifier and got 68.125.
Still, there is a room to improve.

#6

How have you done the feature processing like handling of outliers etc?
Have you implemented the SMOTE technique as well?

#7

I haven’t done any preprocessing with features because all the features are numerical features.