UBC-MDS / forest-fire-area-prediction

This project aims to predict the burned area of forest fires in the northeast region of Portugal, using meteorological and soil moisture data.
https://ubc-mds.github.io/forest-fire-area-prediction/reports/forest_fire_analysis_report.html
MIT License
8 stars 10 forks source link

The grading of Milestone 1 #59

Open Anahita97 opened 2 years ago

Anahita97 commented 2 years ago

Hi Florencia, thanks for your time! We are group 2. I have looked at your comments in the issues, and as you mentioned we have tried to justify why many comments do not apply to our project (as you can see in this issue), however, the TA still seems to disagree with us. The thing is that we understand that if we lost marks if there was an error on our part (and we certainly had a couple of errors), however, the TA has deducted marks for things like addressing class imbalance in a regression problem (whereas we explore the distribution of our target and have considered the log-transformed version of our target to address the skewed distribution). Or, not having a concluding sentence at the end of the EDA document about what variables we are including and excluding, although we explained under every figure why we think we should include/ exclude a predictor, and specifically mention which variables we are including/ excluding in the proposal. Or, she seems to have deducted points for not addressing missing values in the proposal, although we have addressed this in the EDA (as it makes more sense to address this in the EDA as opposed to the proposal). We understand that we did not lose many marks for these, however, this has us concerned about the grading of the next milestones of the project. I really appreciate your time and help! @flor14 @Ivyqiuhan