For the Otto Group Product Classification Challenge, hosted by Kaggle, we have provided a dataset with 93 features for more than 200,000 products. The objective is to build a predictive model which is able to distinguish between our main product categories.
This repository contains a script for the benchmark submission of the competition. Use this script in the following way:
python benchmark.py <path-to-train> <path-to-test> <name-of-submission>
Each argument is optional as the script will guess the right names if you don't change them after downloading and put them in subfolder called data. It will then create a submission called my_submission.csv which should produce the benchmark posted on the leaderboard.
To run the script, you will need to install the following packages:
This script was tested using Python 2.7.9.
We also provide a getting started notebook for those who want to try out neural nets:
Getting started with nolearn/lasagne
If you have a question regarding this script or the competition in general, head to the forum and post them.
