This repo contains my exploratory data analysis and modeling for the Driven Data "Pump it up challenge". The goal was to classify water pumps in Tanzania as functional, functional but in need of repair, or non-functional.
I took this on as an afternoon project one weekend, which yielded a classification rate of .7830 (top 25%). After revisiting the problem, I have been able to boost that to .8252 (just outside of top 1%).
