This is the Thesis project submitted in fullfilment of the Master of Science Degree in Data Analytics at National College of Ireland, Dublin.
The project implemented various machine learning and deep learning algorithms for forecasting electricity consumption for London based households.
-
Exploratory Data Analysis.
-
Data Preprocessing
-
Random Forest Decision Trees (CTREE and RPART) + ARIMA
-
Long Short Term Memory
-
Prophet
-
Gated Recurrent Unit
-
Clustering
Gated Recurrent Unit outperformed all others giving lowest MAPE.