The analysis of trading prices is one of the most challenging tasks in data modelling. For decades, analysts have used statistical and econometric models to try to understand the complex dynamics of security prices generated by globally distributed traders doing business on the world's financial exchanges. In recent years paper based trading, where a trader would fill in an order form and submit it to a broker, has been replaced by electronic trading, resulting in more securities being traded more often. The corresponding increase in the amount of data available for analysis has led many researchers to explore the use of computational intelligence techniques for modelling trading data.
The field of computational intelligence deals mainly with the development of models and algorithms whose structures and mechanisms are inspired by human cognition. In simple terms, computational intelligence seeks to develop models that can reason, understand or learn like a human. The ability to spot patterns, adapt to new and unusual data, and to be robust to non-perfect data are hallmarks of computational intelligence methods. These objectives dovetail very nicely with the requirements of a trader or market analyst - they want to spot trends in past data in the hope they will repeat in the future; they want a model which can adapt to changing market conditions in a controlled, easy to understand fashion; and they want a model which will not completely fail due to potentially noisy data/outliers.
The key aspect of this project is the development of a system that can analyse big data and predict future energy prices. This will be based on the use of machine learning algorithms and ensembles of them.