Kshanti Greene's Web Page

Ph.D. Candidate in Computer Science

University of New Mexico
Advisor: George Luger
Email:
Linked In

Artificial Intelligence Interests:

  • Probabilistic reasoning and graphic models
  • Social and collaborative modeling
  • Networks and distributed behavior
  • Causal learning

Dissertation Topic:

Leveraging consensus and diversity in Bayesian belief aggregation

Existing Bayesian belief aggregation methods utilize opinion pool functions that compute a single value to represent the belief of all contributors. The resulting models break many theoretic assumptions for Bayesian reasoning and they do not represent reality well, especially in cases where there are many diverse opinions. Divergence is a natural result of combining opinions from individuals with different beliefs, backgrounds and experiences. Instead of forming a single consensus value that will average out this diversity, I find clusters of agreement for each probability distribution and propagate the cluster means throughout the network during inference. I use a social network to track the agreement between individuals and utilize graph clustering to find the groups of consensus. I leverage the agreement that occurs across multiple belief estimates to help reduce the complexity that may arise as the means are propagated throughout a belief network. My goal is to enable practical collaborative Bayesian belief modeling from a small group of experts to many individuals, generating both accurate and representative belief networks.

Career:

Publications:

K. Greene, G. Luger. Agreeing to disagree: Leveraging consensus and divergence in Bayesian belief aggregation
To be presented at AAAI 2009 Spring Symposium. UNM Tech Report

K. Greene, T. Goan and E. Creswick. A Collaborative Eye to the Future. In Proceedings of SPIE 2008

K. Greene and M. Hoffman. Coordinating busy agents using a hybrid clustering-auction approach. Presented at AAAI Õ06,
Workshop on Auction Mechanisms for Robot Coordination, Boston, MA.

K. Greene, D. Cooper, A. Buczak, M. Czajkowski, J. Vagle, M. Hofmann. Cognitive agents for sense and respond logistics.
Defence Applications, Multi-Agent Systems. Simon G. Thompson and Robert Ghanea-Hercock, eds. Springer Verlag, 2006.

A. Buczak, K. Greene, D. Cooper, M. Czajkowski, M. Hofmann. A Cognitive architecture optimized for adaptation.
In Proceedings of Artificial Neural Networks in Engineering (ANNIE), 2005.

R. Rommohan, C. Chakrabarti, D. Pless, K, Greene and G. Luger. Logic-Based First-Order Stochastic Language that Learns,
University of New Mexico, department of Computer Science Tech Report: TR-CS-2003-17.

K. Greene, G. Luger, E. Siregar. Model-Combining by Conditioning Rule Sets: Improving on a Signal Classifier for NASA,
UNM Tech Report TR-CS-2003-17.

Education:

Other Interests: