The Artificial Intelligence Society and Dr. Gopal Gupta's research group proudly present: Explainable AI w/ Dr. Gupta!
This is a UTD exclusive event that is sponsored by Professor Gopal Gupta's lab, and challenges participants to apply FOLD-R++ or FOLD-RM algorithm onto Kaggle datasets of the participant's preferred choice. FOLD-R++ is a tool for binary classification whereas FOLD-RM is a tool that is used for multi-category classification. Keep in mind that FOLD-R++ and FOLD-RM only work on tabular data so make sure to filter the myriads of datasets that Kaggle has to an appropriate dataset.
The competition will end on February 20th, at 1 pm (Sunday).
Judging will take place on February 20th, 2-4 pm on teams (click link).
Requirements
1. Analyze the rules generated; see if you can identify any bias/unfairness in the model learned (rules generated represent the model learned).
2. Write a short report on the dataset analyzed and submit it on Kaggle. You can see sample reports done by others on Kaggle. The report should contain a description of the dataset, the rules you generated, analysis of the rules, any bias/unfairness that you can see in the rules, etc. Any other insight you can obtain from the generated ruleset should also be described. You can also write about what new things you learned as well as about your experience.
3. Prepare a video (maximum 5 minutes) documenting your effort and put it on DevPost: https://explainable-ai-dr-gupta.devpost.com. The best way to make the video is to make a few slides, present those slides and record your presentation.
Prizes
Winners
A cash prize worth $100 each for the top four team submissions. If a two people team win, then the prize is split evenly between each team member.
Most datasets
The team that operates on the most Kaggle datasets will receive a cash prize of $100. If the team consists of two people, then there will be an even split of $50 each for each team member.
Devpost Achievements
Submitting to this hackathon could earn you:
Judges
Dr. Gopal Gupta
Prof. Jalal Omer
Huaduo Wang
Parth Padalkar
Judging Criteria
-
How interesting is the dataset
-
How many datasets tried
-
Quality of the analysis of the learned rule-set
Questions? Email the hackathon manager
Tell your friends
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.