View Source

h2. Course Project Topics from Previous Semesters:

h3. Literature Critiques

* Multi-agent data mining: Survey of existing approaches and the potential applicability of an agent-based economic market approach.
* Survey and critical discussion of current technology used for business-to-business Internet markets, and the potential role of agent-based technology in B-to-B Internet markets.
* Critique of selected ACE duopoly (two-firm) studies
* Alternative approaches to game theory (evolutionary vs. classical) in relation to ACE modeling of strategic interaction.
* The importance of increasing returns and path dependence for the development of high technology goods (Microsoft Antitrust Case)
* Critique of selected chapters from Andy Clark's book _Being There: Putting Brain, Body, and World Together Again_
* Critique of studies of eBay and other auction markets using agent-based computational modeling
* Critique of studies on the performance of "Zero-Intelligent (ZI)" traders in double-auction markets. (How important is agent learning in highly structured market contexts?)
* A critique of agent-based software engineering (does agent-oriented programming represent a fundamental advance relative to OOP?)
* A comparative study of two artificial stock market models (Santa Fe ASM and a model due to Doyne Farmer)

h3. Experimental Studies using Originally Developed Software

* An ACE stock market model in which learning agents receiving a stream of actual historical stock market data can condition their investment strategies on fundamental information and/or on data patterns relied on by commonly used technical trading rules (RSI, Bolling Band,...).
* The effect of memory length on the iterated prisoner's dilemma game
* A graphical agent construction kit for the iterated prisoner's dilemma game (generalizable to other games)
* *Hunters: An Agent-Based Artificial Society*, by Andy Schlei.
* A three-player iterated prisoner's dilemma tournament
* Experimental studies using variously extended versions of Chris Cook's Schelling Segregation Model Demo
* Fine-tuning Brian Arthur's El Farol Bar Problem (systematic experiments to explore the robustness of Arthur's key findings)
* Learning in double auctions: A comparison of reinforcement learning with learning represented via genetic programming automata (due to Dan Ashlock)
* The movement of labor and firms in a closed circular-flow economy
* An agent-based model of price-formation in an electronic auction
* Agent-based tools for the forecasting of airline demand
* IPD: How an agent's "conscience" affects the evolution of cooperation
* An artificial stock market model
* An agent-based model for the experimental study of the economic effects of prohibiting smoking in bars