Skip to content

denadai2/energyUva

Repository files navigation

energyUva

This project wants to show the work I've done during my Master's internship in Amsterdam. The focus of this work was finding outliers in the gas consumption of some buildings.

The results are described in this page.

Instructions

  • Clone this pylearn2, which uses the exact pylearn2 version we used.
  • Install the cloned pylearn2

Commands

To train your model you can use

python [PYLEARN_DIR]/pylearn2/scripts/train.py NN-cross_v2.yaml

To predict

python [PYLEARN_DIR]/pylearn2/scripts/mlp/predict_csv.py bests/bestv2-0.pkl test.csv predict.csv -P 'regression'

Releases

No releases published

Packages

No packages published

Languages