Towards self-financing research

Many researchers have suggested using genetic algorithms (GAs) to discover financial trading strategies automatically. This talk will discuss many of the problems involved, and some potential solutions to them. The questions to be answered include: (1) Which securities should be traded, at what time frequency? (2) What should be the form of the trading strategies considered? (3) What historical data should be used to judge trading strategies? (4) What objective function should a trading strategy maximize? (5) Should a GA be used, or some other global optimization method? (6) How can one choose optimal settings for the many parameters of a GA, such as population size and mutation rate? By far the most important question to answer is how one can be confident that a trading strategy discovered using historical data will continue to be profitable in the future. I shall show that this question can be answered successfully using new and old ideas from nonparametric statistics.

Author: Charles Elkan
Computer Science and Engineering
University of California, San Diego

Charles Elkan is an associate professor with tenure in the Department of Computer Science and Engineering at the University of California, San Diego. His main research interests are in artificial intelligence and data mining. Among other work, he has developed motif-finding methods for DNA and protein sequence analysis, algorithms for reasoning about database queries and updates, and methods of formalizing commonsense knowledge about causation. In the field of knowledge-based systems, his paper with A. Hekmatpour entitled "Categorization-Based Diagnostic Problem Solving in the VLSI Design Domain" won a best paper award at the 1993 IEEE Conference on Artificial Intelligence for Applications. In 1998/99 Dr. Elkan was a visiting Associate Professor at Harvard University. Before joining UCSD in 1990, he was a postdoctoral fellow at the University of Toronto. He earned his Ph.D. and M.S. at Cornell University in computer science, and his B.A. at Cambridge University in mathematics.

Towards self-financing research

Click here to start

Table of Contents