Logistic regression solution notebook refactoring

microsoft / ML-For-Beginners

12 weeks, 26 lessons, 52 quizzes, classic Machine Learning for all

https://microsoft.github.io/ML-For-Beginners/

MIT License

68.98k stars 14.24k forks source link

Logistic regression solution notebook refactoring #645

Closed carlotta94c closed 1 year ago

carlotta94c commented 1 year ago

Fixing features encoding (using ordinal encoder for Item size, categorical encoder for all the other features and label encoder for Color) as recommended by scikit-learn documentation
Using F1 score instead of accuracy to evaluate the model since the dataset is highly skewed
Updating plots to better show the data before and after encofings and the relationships between label and features
Fixing ROC curve plot using matplotlib

github-actions[bot] commented 1 year ago

Azure Static Web Apps: Your stage site is ready! Visit it here: https://gray-sand-07a10f403-645.westeurope.1.azurestaticapps.net

bstollnitz commented 1 year ago

Looks great!

jimbobbennett commented 1 year ago

Should we remove the output from this notebook? Would make it smaller and easier to diff

gdequeiroz commented 1 year ago

Should we remove the output from this notebook? Would make it smaller and easier to diff

@jimbobbennett Do you mean just for the PR or in general? Since this is a solution notebook, we would like to have the output (+ we are trying to be consistent with the other solution notebooks from the other sections.

jimbobbennett commented 1 year ago

Should we remove the output from this notebook? Would make it smaller and easier to diff

@jimbobbennett Do you mean just for the PR or in general? Since this is a solution notebook, we would like to have the output (+ we are trying to be consistent with the other solution notebooks from the other sections.

I was thinking in general, just this notebook was in the PR so made me think of it. Is it intentional to have the output in the notebooks then?

gdequeiroz commented 1 year ago

Should we remove the output from this notebook? Would make it smaller and easier to diff

@jimbobbennett Do you mean just for the PR or in general? Since this is a solution notebook, we would like to have the output (+ we are trying to be consistent with the other solution notebooks from the other sections.

I was thinking in general, just this notebook was in the PR so made me think of it. Is it intentional to have the output in the notebooks then?

Yes! :)

jimbobbennett commented 1 year ago

Can't add this inline as the file is too large, but in the Build your model section, this line:

from sklearn.model_selection import train_test_split

is in both of the first 2 cells, it doesn't need to be in the second one.

github-actions[bot] commented 1 year ago

Azure Static Web Apps: Your stage site is ready! Visit it here: https://gray-sand-07a10f403-645.westeurope.1.azurestaticapps.net