Discuss the homework here!
Here is the homework submission form: https://forms.gle/6rRzJDVFb9dTAtCS8
If you’d like to get a certificate for this workshop, you must attempt the quiz questions in the Google Form and upload your homework solution notebook (the extension of which must be .ipynb).
@ionizedbeing We appreciate your enthusiasm. The homework submission link is active now and you can find it at the top of this thread here.
Please submit your notebook on the submission link. We hope to see your submission soon!
Even after One-hot encoding the position feature, the model’s r2 score didn’t improve much. Is this because B7 (coefficient of position feature) is very small (-0.66 to 0.90) as compared to B6 (coefficient of big club) which is very big (8.58) and hence has a greater impact on the prediction?
@m0sh1 Yes R2 score will not change much, but its important from interpretation stand point of view. The categories, even though they are represented by numbers are essentially categories. For example, 3 is for defenders, 4 is for goalkeepers, here 3 < 4 does not make sense since these are both categories. Also, since there are multiple predictors, the impact of one hot encoding will not necessarily have a lot of impact on the R2 score, but will have correct interpretations for each category. Yes, the impact of big_club is significant on the predictions and it makes sense from interpretation point, if a player is from one of the Top 6 clubs, they will most likely have higher market value. But remember, for simplicity, we haven’t used any Standardization methodologies for this homework and Standardization might have some impact on the coefficients.