The Nomination for Prediktering av fiendeintention, baserat på Bayesiansk hypotesprövning by Fredrik Johansson

In this thesis, the general problem of how to find appropriate prior distributions for Bayesian networks, has been addressed by developing a tool for data collection. By integrating a Bayesian network into the simulation framework GTSIM (Warston & Persson, 2004), different scenarios where an enemy's intent is predicted in terms of probabilities for different courses of action can be visualized and evaluated. Such a tool may in the long run facilitate the encoding of expert knowledge into conditional probability tables and thereby enable predictions to be more automated.

This thesis is of high quality and is well written. It is also a good example of how AI techniques can be used in real world applications (this work has been conducted in cooperation with Ericsson Microwave Systems, Göteborg, on the one hand, and Markstridsskolan, Skövde, on the other). Fredrik has recently submitted a paper (Johansson & Falkman, under submission) based on the thesis to the 9th International Conference on Information Fusion (www.fusion2006.org).

Fredrik is now a PhD student in Skövde's Information Fusion research program, and continues to work on closely related issues in cooperation with Ericsson Microwave Systems.

Mikael Johannesson

References:

Johansson, F., and Falkman, G. (2006). "Implementation and integration of a Bayesian Network for prediction of tactical intention into a ground target simulator", Submitted to The 9th International Conference on Information Fusion (www.fusion2006.org).
Warston, H., and Persson, H. (2004). "Ground surveillance and fusion of ground target sensor data in a Network Based Defense", In Proceedings of the 7th International Conference on Information Fusion, pp. 1195-1201.