Our project centers around designing an automated negotiating agent for submission to the Supply Chain Management League (SCML) division of the ANAC2020 competition. Our agent, Monty Hall, can be divided into three components -- a business planner, a negotiator, and a contract signer. The business planner uses a model of the market uncertainty to determine ideal quantities to buy and sell at each step, solved for using a linear program. The output gotten from the business planner is fed into the pre-negotiation agenda of the negotiator. Our negotiator makes use of an aspiration-based strategy. We further devised a marginal utility calculator, which finds the added expected utility of each new contract with a linear program, but this was not included in the agent. Negotiated contracts are given to the contract signer, which is able to select which contracts we would like to sign and which we will discard, since we only have a limited amount of inventory. This is done by modelling the signing problem as a knapsack problem and then using the dynamic programming approach to solve it, which lets us filter out which sell contracts we want to sign. We also took an approach using a linear program which modeled the trust we have that other agents will sign, but this was not included in the submission. Preliminary tournament results show Monty has a median of zero but the highest mean of all agents.
Li, Edward, Tsatsaros, James, Silverston, Daniel, et al.,
"Automated Negotiating Agent For Supply Chain Management League"
(2020).
Summer Research Symposium.
Brown Digital Repository. Brown University Library.
https://doi.org/10.26300/js87-b326
Each year, Brown University showcases the research of its undergraduates at the Summer Research Symposium. More than half of the student-researchers are UTRA recipients, while others receive funding from a variety of Brown-administered and national programs and fellowships and go …
