List of publications

Current Projects:
Computing using soliton collisions
In most present-day conceptions of a "computer," information travels between logical elements fixed in space. This is true for abstract models like Turing machines, as well as real silicon-chip-based electronic computers. This work views computation in an entirely different way: information is carried through space by particles, computation occurs when these particles collide.

The most recent work focuses on the storage of information in states of optical solitons, the transformation of these states when solitons collide, and the computational power of solitonic systems. The long-term goal is the construction of a new kind of computer using soliton collisions, as well as the basic understanding of information processing in nonlinear waves. The most recent work is reflected in the following papers.

"Particle Dreams" #5, Matt Howarth, Fantagraphics Books, 1987.

Solitons and Particles in Cellular Automata: a Bibliography

Embedded Computation using Solitons and Particles: a Bibliography
Auction Theory and Evolutionary Biology

A paper with Prof. John Morgan, Haas School of Business, University of California at Berkeley, on the formulation of the replicator equation, and how spiteful behavior may arise in evolutionary biology. Comments most welcome.
A paper on spiteful behavior in auctions, with John Morgan and George Reis, '01. Comments most welcome.

Agent-based simulation

Market-based models have important applications in distributed resource allocation and are becoming increasingly important as agent technology emerges in the internet. In situations with many complex heterogeneous agents, simulation is the only tool for gaining insight into the properties of systems short of actual implementation. This work focuses on the simulation of computational multiagent markets, sometimes called robot markets.

The following papers show that trend-based traders can de-stabilize robot markets, causing both inflationary and deflationary price bubbles.
The properties of online auctions reflect human behavior. The following is a start at modeling such behavior and its effects in auctions with deadlines.

Robot Markets: a Bibliography