Joseph Reisinger
This summer Google Research has graciously agreed to host my work on semi-supervised latent variable models.
Current projects include: Nonparametric Bayesian Approaches to Semantification/Information Extraction, Leveraging Evolvability in Search, and Evolutionary Games with Unknown Payoff Distributions.
Natural Language Processing
Multi-Prototype Vector-Space Models of Word Meaning , with Ray Mooney. NAACL-2010.
Latent
Variable Models of Concept-Attribute
Attachment
, with Marius
Pasca. ACL-2009.
(supplemental
material)
Bootstrapped
Extraction of Class Attributes,
with Marius
Pasca. WWW-2009.
Weakly-Supervised
Acquisition of Labeled Class Instances using Graph Random
Walks.
Partha Pratim Talukdar,
Joseph Reisinger, Marius
Pasca, Deepak
Ravichandran, Rahul
Bhagat and Fernando
Pereira. EMNLP-2008.
Machine Learning
Spherical Topic Models, with Austin Waters, Bryan Silverthorn and Ray Mooney. NIPS-2009 Applications of Topic Models Workshop.
Online Kernel Selection for Bayesian Reinforcement Learning, with Peter Stone and Risto Miikkulainen. ICML-2008. (video)Coevolving
Strategies for General Game Playing
with Erkin Bahceci,
Igor Karpov and Risto Miikkulainen. IEEE-CIG-2007.
Evolvability
Acquiring
Evolvability through Adaptive Representations, with
Risto Miikkulainen.
GECCO-2007.Best Paper Award
Selecting for
Evolvable Representations
with Risto Miikkulainen. GECCO-2006.
Towards an
Empirical Measure of Evolvability
with Kenneth Owen
Stanley and Risto Miikkulainen. GECCO-2005.
Evolving Reusable Neural Modules, with Kenneth O. Stanley and Risto Miikkulainen. GECCO-2004.