The following feedback is based on the release v0.2.0.
If you have already updated your repo respecting some of my feedback in future versions> v0.2.0, feel free to ignore the corresponding feedback on v0.2.0.
Documentation: Is the project properly documented?
Yes, it is great.
Code:
The code is readable.
One optional improvement would be:
For download_data.csv, you can write some error handling code for downloading data. Our group was deducted points because we didn't do it, although it's not required in the written requirement.
Analysis and reasoning: Is the analysis correct?
Yes, it is great.
One improvement would always include a baseline model for model comparison. No appearance of the baseline model is mentioned in your report. We don't know if the models perform better than without a complex model or not.
Communication:
Yes, it's easy to follow and the report communicates the findings effectively. Good job!
Thanks you @KangboLu for the insightful comments, they are really helpful!
We will discuss among our group to decide whether we include the baseline model.
The following feedback is based on the release v0.2.0.
If you have already updated your repo respecting some of my feedback in future versions> v0.2.0, feel free to ignore the corresponding feedback on v0.2.0.
Yes, it is great. One improvement would always include a baseline model for model comparison. No appearance of the baseline model is mentioned in your report. We don't know if the models perform better than without a complex model or not.