I hereby nominate Wah-Sui Almberg for the SAIS Best AI Master's Thesis 2002 award. His master thesis "Improved Pricing on the Stock Market with Trading Agents" was successfully defended at Stockholm University, Department of Computer & Systems Sciences (DSV), in the spring of 2002.
I supervised his work, and Almberg also benefited from having an active external supervisor at OM: David Lybäck. Almberg became part of a project group including myself, Lybäck, and several other senior researchers, which proved helpful in many ways. After the completion of his master, I have kept in contact with Almberg, and I have recently agreed to supervise also his Ph.D. thesis. Given the complex topic, Almberg's background in physics (he has double degrees) proved useful. In my opinion, only one student in 100 could successfully grasp the subtleties of financial mathematics and at the same time do theoretical computer science.
The main reason for my nomination of Almberg, however, is that I think that the artificial intelligence aspects of his work is truly innovative. For the history-based reasoning of his agents, Almberg devised his own version of the second derivative. Not only did he defend this theoretically, but he himself implemented this reasoning and found time within his master for a simulation study too.
Finally, I find his style of presentation charming and engaging, and like most of his work: unique.
Magnus Boman ("docent" at DSV, leader of the HUMLE lab at SICS)