Short-Term-Residential-Load-forecasting-using-LSTM

This is the unofficial implementation of the paper https://www.researchgate.net/publication/319886687_Short-Term_Residential_Load_Forecasting_based_on_LSTM_Recurrent_Neural_Network. The paper introduces SGSC dataset https://data.gov.au/dataset/ds-dga-4e21dea3-9b87-4610-94c7-15a8a77907ef/details which is a very comprehensive dataset containing customer's electricity consumption patterns along with other data. A subset of data is used in the paper which is also included in this repository. Main file creates and tests the LSTM models on the data of differnet customers and compares the performance of MAPE and MAE as loss function.