This is the first of AI@IFIC hackdays, this time focusing on an issue important to climate change, the accurate prediction of variable energy sources such as wind and solar.
During this half-day we will explore the use of Machine Learning techniques to forecast variable low-carbon energy sources. We will start by using public sources such as
and correlate them with weather measurements and projections to predict future production.
The idea is to explore a range of methods such as RNNs and LSTMs as well as regular NNs.
Previous to the meeting, we ask participants to register in Google Colab and/or setup their computers to run jupyter notebooks and conda. We may release a docker for this meeting.