Live Machine Learning Demo for Energy Markets

Using an LSTM network (a type of recurrent neural network or RNN) for reinforcement learning to forecast time series in UK energy, specifically month ahead UK electricity prices

The recurrent neural network deployed here learns by targeting a goal (minimising the difference between forecast and actual price) then rewarding / penalising action taken towards achieving the goal (attributing weights to neurons in a network and iterating through layers - so called "deep learning").


The rewards and penalties are the RNNs own internal feedback system and are computed each day - check back for updates!

Plot 5

Our prototype uses Binder, GitHub and a Jupyter Notebook with Python code using Keras / TensorFlow.



Click below to run to a live forecast with real time weather, energy and FX data.

Real-time ML app
Energy Chatbot
Daily Macro NLU
AI Blog
CE AI services
AI - exec guide

Please submit your details below if you would like to find out more about ChangingEnergy AI services

You can also provide comments below or feedback