Kaggle Final Project

The final project was a Kaggle competition to predict standardized test essay grades. Although I still had lots of ideas, when I wrapped up a week early I was in first on the public leaderboard, and maintained that to the end. After it was over the private results gave first to Maura, who implemented some awesome ensembling. For commentary take a look at Rachel’s blog post. There’s a bit of discussion in the forum, including my write up of my code.

Visualization

During the competition I maintained a visualization of the leaderboard, which shows everyone’s best scores at that moment. Will Cukierski at Kaggle appreciated it, and apparently the collective impetus of Rachel and I encouraged them to make a competition out of visualizing the leaderboard! See Rachel’s blog post about it for some more info (and a nice write up about my mistakes).

Now back to studying for finals…

The final project, accounting for most of the grade, is a Kaggle competition. It’s based on an earlier competition, and the goal is to develop a model to grade standardized test essays (approximately middle school level). As an auditor Rachel asked me not to submit, but my cross-validation suggests my model (linear regressions with some neat NLP derived features) is still besting the public leaderboard (Quadratic Weighted Kappa Error Measure of .75), but who knows.

I thought I could easily adjust my MITRE competition leaderboard graph to Kaggle’s CSV, and it turned out to be pretty easy. The biggest issue ended up being that MITRE scored 0 to 100, and this scores 0 to 1. That had some unintended consequences. launchd + python + R should upload this every hour or so (when my laptop is running).

I’m frankly surprised Kaggle hasn’t done something like this before. Maybe if I have a bored evening I’ll try to do it in D3, which should look much nicer.

Update, 12/16: I’ve posted a followup after the end of the semester.