First submission (python)

Prerequisites

• python 2.7
• numpy
• scipy
• pandas
• sklearn
• ipython

Baseline submission

The file baseline.ipynb provides simple Gradient Boosting training on the whole training data set. The full preevaluation procedure is demonstrated here: CvM and KS are calculated on check datasets, weighted ROC AUC is calculated on the training data set (for details about evaluation see Evaluation procedure).

NOTE

To provide agreement we train classifier on the subset of all available features.

First submission steps

1. Clone the repository

   git clone https://github.com/yandexdataschool/flavours-of-physics-start

2. Follow to the instructions in the tau_data/README.md

3. Run all cells in the notebook baseline.ipynb, this step will generate a baseline.csv, the file prepared to submit