Michael Kampouridis

I am interested in the use of Computational Intelligence (CI) techniques in business-related problems. So far, I have used Evolutionary Algorithm and Artificial Neural Network techniques in the fields of Finance, Economics, and Business.

Current projects

My current research can be divided into two areas: algorithmic trading, and weather derivatives pricing.

With regards to algorithmic trading and financial forecasting in general, now, in the aftermath of a global financial crisis, it is more important than ever to have a better understanding of the markets and be able to forecast their movement. Directional changes is a new concept, which is based on the idea that an event-based system can capture significant points in price movements that the traditional physical time methods cannot. I am currently using evolutionary algorithms to create trading rules, by taking advantage of this new concept.

Another area I am heavily involved with is weather derivatives. Weather derivatives are financial instruments used as part of a risk management strategy to reduce risk associated with adverse or unexpected weather conditions. My aim is to develop a model of pricing weather derivative contracts by the use of genetic programming methods. Until now, there is no generally accepted framework for pricing such derivatives, as it happens with other (non-weather) derivatives (i.e. Black-Scholes model). This is a major problem, as it leads to incorrect predictions of the contract prices, thus resulting to significant financial loses. On the other hand, developing a novel genetic programming algorithm to create a generic pricing framework has the potential of great impact in the sector, by solving a problem that has existed since the introduction of weather derivatives in 1997, much like the Black-Scholes model did for options pricing in 1973. This is a significant problem, which is getting more and more attention by both industry and academia.

Past projects

I have used Genetic Programming to develop a financial forecasting tool, named EDDIE, which I used for predicting buy opportunities based on data from daily closing prices. I am deeply interested in the field of financial forecasting, and I continuously look for new search methods that can improve the predictability of forecasting tools such as EDDIE.

In addition, I have developed a financial model to study market dynamics and market behaviour by using Genetic Programming, along with Self-Organizing Maps. Financial modeling is another area that interests me and I believe that it has much to offer in understanding the financial markets and the decision-making process.

Furthermore, I have applied Genetic Algorithms and other heuristic techniques to a Telecommunications problem, where I built an intelligent system alongside an economic model, for British Telecoms (BT). This system acts as a decision support tool for the investment of Fibre Optic Networks, by advising BT on the optimal time and location for deploying a network.

Lastly, I have also been working on Automated Bargaining for Price-Speed (P-S) Optimised Negotiation. While Price-only optimisation is very popular and well-known, Price-Speed optimisation is relatively new. I am the first to bring GP into P-S optimisation and results have already demonstrated the GP’s superiority against other state-of-the-art algorithms.

My interests are not, however, limited only in the above applications. I am always open in any type of interdisciplinary research that includes the use of Computational Intelligence. I am thus very keen on bringing CI into business, economics, and finance.