In this project, I design and implement a negotiation agent on GENIUS[1] for multiparty negotiation, which is a multiagent system developed at the University of Delft, The Netherlands. The implementation is made in Java by using GENIUS negotiation environment. The negotiation agent is designed to be used in the “Alternative Offer Protocol (AOP)”.

Appaloosa Agent is a time based conceder, which adjusts according to its minimum accepted utility threshold, which depends on a model constructed by logistic regression. Since machine learning libraries are allowed, I used Weka library, which is a machine learning library for Java, in order to build a logistic regression model. You can see the the details from "Appaloosa" PDF. GroupY qualified as one of the finalists in PRIANAC 2018 (Pacific Rim International Automated Negotiation Agents Competition) [2].