TII_Wide&CNN_Electricity_Theft_Detection

This is the source code of our paper on electricity-theft detection published in TII in the 2017 year.

• Zibin Zheng , Yatao Yang , Xiangdong Niu , Hong-Ning Dai, Yuren Zhou. Wide & Deep Convolutional Neural Networks for Electricity-Theft Detection to Secure Smart Grids[J]. IEEE Transactions on Industrial Informatics, 2017:1-1.

Source Code Introduction

• function.py : include custom data processing functions that are needed in the experiment.

• keras_metric.py : include the AUC, MAP@100, and MAP@200 metric function which will be executed on each epoch.

• wide_cnn.py: the source code of our Wide&CNN model.

• run.py : the main file, we can run to get the experimental results.

• log/ : store experimental result logs.

• data/ : store experimental datasets.

Dataset Download Address

LINK: https://pan.baidu.com/s/17exE465yp79HWJ06qRvWKg Extraction code: i6xy

It should be noted that tha data contains two files.

• after_preprocess_data.csv : it is the electricity consumption data of users after data preprocessing.

• label.csv : it is the label data of whether a user steal electricity. Each line corresponds to each user.