Prof. John Moody Computational Finance Lab International Computer Science Institute Berkeley & Portland

Professor Moody is on leave from full-time academic work to run J E Moody & Company LLC, a quantitative hedge fund management company. If you would like to inquire about employment opportunities, please contact the company.

Machine Learning in Finance, December 9, 2005, Whistler, BC.

Research Interests

Professor Moody does research in data mining, machine learning, time series analysis and computational finance. His recent research includes work on direct reinforcement methods for active trial and error learning, regularization and priors for supervised learning and applications to finance and time series analysis.

Highlights of his previous research in machine learning and neural computation include networks of radial basis functions, stochastic learning algorithms and theory, and his proposal of the effective number of parameters. Moody's research in time series analysis has included work in wavelet denoising, nonlinear forecasting, and methods for reducing prediction risk.

Selected Publications

Selected papers on Direct Reinforcement Learning are available for download:

Stochastic Direct Reinforcement: Application to Simple Games with Recurrence. John Moody, Yufeng Liu, Matthew Saffell and Kyoungju Youn. Artificial Multiagent Learning, Sean Luke et al. editors, AAAI Press, Menlo Park, 2004. (PDF)

Learning to Trade via Direct Reinforcement. John Moody and Matthew Saffell. IEEE Transactions on Neural Networks, Volume 12, Number 4, Pages 875-889, July 2001. (PDF)

Performance Functions and Reinforcement Learning for Trading Systems and Portfolios. John Moody, Lizhong Wu, Yuansong Liao & Matthew Saffell. Journal of Forecasting, Volume 17, Pages 441-470, 1998. (PDF)

A full list of publications is available here .

Professional Activities

Moody is a member of the Editorial Boards of Quantitative Finance and Neural Processing Letters, the Board of Directors of the Neural Information Processing Systems (NIPS) Foundation and a member of the IEEE Technical Committee on Computational Finance.

Moody served as Program Co-Chair of the IEEE conference Computational Intelligence in Financial Engineering (CIFEr*2003) that convened in Hong Kong in March 2003. He was Program Co-Chair of Computational Finance 2000 (London Business School, May 2000), and is a past General Chair and Program Chair of the Neural Information Processing Systems (NIPS) conferences.

Teaching and Education

Prof. Moody has taught graduate courses on a wide range of topics as a faculty member in Computer Science at Yale University and OGI. Courses in computer science and related areas have included neural computation, machine learning, statistical learning, computational neuroscience and numerical methods. Topics in computational finance have included options and futures, financial time series and financial markets and trading.

Prof. Moody founded OGI's Computational Finance Program in early 1996, and served as its Director until June 2002. This highly-regarded program was among the first MS programs launched in quantitative finance, and was the world's only such program based in a computer science department. Moody's success in building OGI's Master of Science in Computational Finance earned him the first ever OGI Faculty Award for Educational Contribution in 2001.

Appointments and Degrees

Moody joined Berkeley's ICSI on September 1, 2003. He was previously a Professor in Computer Science & Electrical Engineering at the OGI School of Science and Engineering in Portland, Oregon. Prior to that, he held positions in the Computer Science Department and Neuroscience Program at Yale University and at the Institute for Theoretical Physics at the University of California at Santa Barbara. He received his B.A. in Physics at The University of Chicago, and his M.A. and Ph.D. in Theoretical Physics from Princeton University. While studying at Princeton, he held a Hertz Foundation Fellowship.

Contact Info