Research
Artificial Intelligence
Comparing Interference And Facilitation Models. A short comparison between two models attempting to explain both the Picture-Word-Interference effect and the Stroop effect.Prisoners Dilemma (with Bastiaan Daems). Comparative research on different strategies for the classic game-theory prisoners dilemma.
Trading Agents (with Bastiaan Daems). Experiment report on a trading-agent implementation with focus on agent language and communication.
Multiple Output Echo State Networks (with Arjan Eising). Experiment report on an echo state neural network designed to yield multiple different output forms.
SSL/TLS Session-Aware User Authentication. A short review of this paper by R. Oppliger, R. Hauser and D. Basin discusses the proposed solution and notes possible obstacles and alternatives.
Solving the partition problem using a genetic algorithm. A method is reviewed where genetic algorithms are used to incrementally approach a possible solution to a classic problem: given an array of numbers, find a partitioning such that two heaps of equal size emerge. Using brute-force techniques, such a problem would mean computing the sizes of each heap for 2N possible combinations. While genetic algorithms may not always be able to ascertain the best solution has been found, they can generally find a good solution relatively quickly.
Particle Filter Optimalisation (with Erick Wilts). Experiment report on a localisation algorithm using a particle filter. Through parameter sweeps optimal deviation values are derived resulting in the succesfull localisation of a Pioneer robot.
Vector Based Path Planning (With Erick Wilts). Experiment report on a vector based path-planning algorithm. Using image and infrared distance sensors an Aibo robot obtains a list of attractors and repellors cumulating in an optimal path towards its goal.