Same comment around notes in your notebook remains, please address that. Seek inspiration from the notebooks I and Cyril have shared.
In your multivariate analysis, you need to include the dependent variable to understand the relationship.
In your code you mention you are downsampling but there is no implementation of this. Moreover, you need to justify the reason for balancing the data, you need to have this visualized under the EDA section.
Please prepare a confusion matrix similar to the one in the notebook to show other classification metrics.
Now that you have covered week 4 content, it would be great if you tuned the model.