Shelter Animal Outcomes
This kaggle competition utilizes a dataset of intake information from the Austin Animal Center to predict the outcome for each animal delivered to the shelter.
Possible outcomes are: Adoption, Died, Euthanasia, Return to owner, Transfer.
The jupyter notebooks in this repository contain exploratory analysis as well as a gradient boosting machine to make outcome predictions. This model is optimized to reduce the multiclass logarithmic loss.
Since the data size is relatively small, training was done on a late 2013 Macbook Pro.
The dates of this kaggle competition are March 21, 2016 through Jul 31, 2016. There are no points or tiers awarded.