Bannière Faculté des sciences DIC
Title : SÉMINAIRE: Network-based and Survival Analysis Approaches for Electricity Load Forecasting
Tutor : Étienne Gael Tajeuna
Number : 39/18
Status : Not exceeded
Begin : Thursday, 29 November, 2018 à 10:30AM
Location : Salle PK-5115, Pavillon Président-Kennedy (PK), 201, avenue du Président-Kennedy, H2X 3Y7
Bookable : 12


In the field of energy analysis, time series forecasting techniques are widely used to predict customer electricity consumptions. To enhance the electricity forecasting accuracy, in current approaches, clustering techniques are first applied to identify groups of customers exhibiting the same electricity load profile, from which a representative consumption pattern can be extracted. This pattern is later used to predict customers’ subsequent electricity consumption. In the vast majority of clustering approaches, authors use the entire data set as input to identify customer consumption groups. However, using such a global methodology may affect load forecasting accuracy, as it does not consider the fact that customers’ consumption behavior may change at any time. Predicting customers' electricity consumption with such unstable behavior poses a serious problem for existing models. To overcome this constraint, we propose a hybrid approach capable of handling customers’ changeable electricity consumption. In summary, we propose a network-based approach which involves tracking cluster structures over time. From the evolving clusters, we develop a principled framework which simultaneously utilizes long-short term memory (LSTM) recurrent neural network and survival analysis techniques to forecast electricity consumption.


Etienne Tajeuna received his BSc degree in 2011 from the University of Yaounde I, Yaounde, Cameroon. He had a professional MSc with the University of Jules Verne Picardie, Paris, France in 2012. In 2015, he got his MSc from Université de Sherbrooke, Quebec, Canada. He is currently a PhD candidate at Université de Sherbrooke. His research projects include time varying social and information networks.

Document PDF : Network-based and Survival Analysis Approaches for Electricity Load Forecasting