Open poojapathak1725 opened 1 year ago
Maggiebgg
commented
1 year ago Hello thank you so much for your feedback! I am a little confused with “make up these points”, does it mean we can get all the points back that were deducted in this assignment if we do well on the check point?
poojapathak1725
commented
1 year ago Hi, yes if you make the necessary changes by the Checkpoint and incorporate this feedback, I will grant you back the points you lost on this proposal.
I was intentionally harsh in my grading here - but the main goal is to steer your group in the right direction and ensure you're on the right track to earning a high score on the final project :)
Hope this helps! Let me know if there's anything else I can help with.
~Project Proposal Grade: 5.75/9~
Updated Project Proposal Grade: 9/9
Title and Abstract (0.5/1)- Add an informative title to your project. Which algorithms will you be using- mention them. How do you plan to measure/quantify the popularity of a new video game? The last sentence of this section mentions that you will be predicting a new game - are you predicting a new game or predicting the popularity status of a new game? (-0.5)
Background (0.5/1) - Good start. Make sure to link the prior work cited "Predicting the popularity of Games on Steam." (-0.5) Include a line on the kind of predictive model used, metrics used, and the result of this prior work. Is this your benchmark model? Mention clearly. Seems like you missed the last sentence of this section.
Research Problem Statement (& Significance, Purpose of Project) (1/1.5)
Could rephrase more clearly - "machine-learning based technique" is vague. I would include a line that clarifies how you plan to quantify and measure the popularity of a video game. (-0.5)
Data - (1/1.5) Good start - rewrite this section in bullet points for better legibility. What kinds of feature engineering/feature selection techniques will be used? Do any features need to be one-hot-encoded? How will you deal with class imbalance issues? Mention these points (-0.5).
Proposed Solution (0.5/1.5)
I'm confused with this section- what two algorithms are you combining to create the binary classification model?
I think doing correlation analysis on some features to analyze the impact on the target feature is great - could also include statistical tests such as Pearson's/correlation heatmaps.
I'm not 100% clear on the second algorithm you will be combining into the binary classification model. Additionally, we are looking to compare/contrast the performance of at least three models based on chosen metrics along with cross-val/grid search etc. so consider looking into additional models as well.
Please reply to this issue clarifying your steps to build this model. Some details missing - Train/Test splits, cross-validation techniques, hyperparameter tuning, etc. (-1)
Metrics (1.5/1.5) OK. I would suggest analyzing the confusion matrix as a whole as well, Precision, Recall, and then the F-1 Score to give a comprehensive overview of the performance of your model.
Ethics and Privacy (0.5/0.5) OK. Keep adding to this section as you continue your analysis.
Team Expectations (0.25/0.25) OK.
Timeline (0/0.25) Not changed for the team. (-0.25)
Other Comments - Good start. There's a lot you can do with video game data. Make sure to reply to this issue with the clarification and consider the feedback for the upcoming checkpoint as you can make up these points.
You can reply to this feedback below. Contact me anytime if you want help improving your project or have any questions at all!