For #2 : Logistic Regresssion on winequaliy.csv

[SanchiMittal]

I have used Logistic Regression classifier to perform binary classification on winequality.csv and classify the test data into recommended or not recommended wine.
@dzeber and @mlopatka Please let me know what improvements I need to make.
Thanks

[SanchiMittal]

I will add other ML models like KNN, SVM, Naive Bayes etc. in the next step and present a comparative study. @dzeber @mlopatka Kindly review my work till now, guiding if I need some changes in my workflow.
Regards.

[dzeber]

Thanks for this PR. The notebook is well-documented and easy to follow, as are your modules. This is sufficient to satisfy the startup task #2, but you're welcome to dig into the modeling further.

Please add a comment indicating how you decided on this train-test split ratio. We are asking for this input at this point because one of goals of the project is to better understand how this choice influences the outcomes of the model.

Also, do you have any other thoughts on the results from your classification report and confusion matrix? Notice that the model seems to assign most wines to the non-recommended category regardless, which might be inflating the overall accuracy. It would be interesting to see the results on an undersampled training set.

[dzeber]

I would not both printing this since you are also displaying the visualizations.

[SanchiMittal]

Thank you for the review. I will implement the above suggestios in my work. Also, I would like to dig more into modelling and study the performance of different models.

[mlopatka]

Excellent work! Thank you for addressing earlier feedback.