Saving the world in which we are living is more than ever important in order to avoid a situation without a possible way back. Although we are in a critical situation, it is still possible to move towards a sustainable world and help individuals to adopt sustainable behaviors in terms of energy consumption.
In order to help individuals, we must know their current level of sustainable behavior.
Although there exist various theoretical models to reach this goal, there is no technical solution. This research proposal aims at filling this gap and providing developers with an innovative framework that will enable to predict their level of sustainable behavior. This framework will include a personalized machine learning algorithm or an ensemble approach combining several machine learning algorithms (e.g., clustering, neural networks, decision trees, support vector machines, etc...).
To validate our framework, we plan to use a rich dataset called SHEDS that contains over 1'200'000 data points that are focused on energy consumption in three domains: electricity, heating and mobility.

• Date début : 01.01.2019
• Date fin : 31.07.2019
• Montant: 49 808 CHF
• Financement : Hasler Foundation Projects