KAGGLE MINI COMPETITION: IMAGE CLASSIFICATION TASK
This is the Kaggle mini-competition, part of CSCI 4622  that you will be participating. In this competition, you will work with a dataset to classify German traffic signs. This is a simple clean dataset in which each image is 32x32x3 sized which is flattened for convenience purpose to 3072 features. Training and validation data are available in the 'Data' tab.
Remember:
You should be working in teams of 2-3 people. You are not required to physically code together but you are encouraged to try different things and select the best model you'd like to submit.
Competition is open from Friday (4/10/2020) at 2 PM to Friday (4/17/2020) at Midnight.
Your grade on all the mini competitions account for 20% of your total grade
How does this work?
Your goal in this exercise is to build the best model, as a team, to predict the image's class in this dataset (more information in the Data tab). How you code and work as a team is up to you: you can all work on the same code and divide work, or build different models and ensemble amongst them, or any other method you prefer. Once you build the model, you need to upload your submissions. You will have access to labeled training data and unlabeled validation data. Your model will be evaluated based on the log-loss obtained on this validation data.
Based on your log-loss, you will be assigned a rank on the Kaggle Leaderboard. This way you are able to monitor your standing in this competition.  Once the competition is closed, the person at the top of the leaderboard is the winner. But wait, there's a twist. We will hold out an additional subset of the data that your model will be evaluated on. You will be unable to see the scores on this set until the competition closes. This means rankings can change when the competition closes. Therefore, though you try to improve your rankings, remember to keep your model generalizable so that you can have the last laugh.
Deliverables a.k.a what you will be graded on
Your code(s)
Short one or two-page report summarizing everything you did to build your best model. It is recommended that you limit it to one page unless there is something you really want to mention that can't be fit. (You must submit a report in order to get a grade)
Individual's contibutions in brief
Update Regarding Data and Timeline
There was a problem with how labels were specified in the data. They are now one-hot encoded. If you don't have access to these one-hot encoded data please re-download.
The timeline has now been updated, the competition will run through midnight on
Friday (4/17/2020)
and your reports are due by midnight on
Monday (4/20/2020)