Dissertation Related Publications

Karl Pfleger, June 2002, On-Line Learning of Predictive Compositional Hierarchies, Ph.D. dissertation, Computer Science Department, Stanford University, Stanford CA.

The next three items are smaller, denser versions of portions of the dissertation.

Karl Pfleger, 2004, On-Line Cumulative Learning of Hierarchical Sparse n-grams (PDF version). To appear in ICDL-04. [This paper covers, in conference paper length, the n-gram section of my dissertation, chapters 7-8. Read those chapters for the more detailed version.]
Karl Pfleger, 2004, On-Line Learning of Predictive Compositional Hierarchies by Hebbian Chunking (PDF version). In the 2004 AAAI Fall Symposium: Compositional Connectionism in Cognitive Science. [This paper very briefly (in 4 pages) covers the connectionist model of my dissertation, primarily material from chapters 4 and 6. Read chapters 4, 5, and 6 for more details.]
Karl Pfleger, 2000, Learning Predictive Compositional Hierarchies, Technical Report, KSL-01-09, Computer Science Department, Stanford University. [This paper is a manifesto of sorts for my dissertation work, covering motivations and much of the related work. Read the dissertation's introductory chapters (1-3) and related work chapters (9-10) for more detailed treatment.]

The dissertation itself supercedes everything else below here.

Karl Pfleger, 2000, Scaling Up Grounded Representations Hierarchically, presented at the AAAI-2001 Spring Symposium on Learning Grounded Representations. (See Learning Predictive Compositional Hierarchies tech report above.)
Karl Pfleger, 2000, Learning Predictive Compositional Hierarchies, presented at the AAAI-2000 workshop on New Research Problems for Machine Learning. (See tech report of same name above.)
Karl Pfleger, 2000, Partial N-grams. Presented at the ICML-2000 Workshop on Machine Learning from Sequential and Temporal Data. (See On-Line Cumulative Learning of Hierarchical Sparse n-grams paper above.)
Karl Pfleger, 1999, Learning Compositional Hierarchies by Data-Driven Chunking (1-page abstract), in the Proceedings of the Sixteenth National Conference on Artificial Intelligence.
Karl Pfleger, 1997. Learning of Compositional Hierarchies for the modeling of context effects. KSL Technical Report KSL-98-04, Computer Science Department, Stanford University.
Karl Pfleger, 1997. Context effects and learning of hierarchical compositional structure (1-page abstract), in the Proceedings of the Nineteenth Annual Conference of the Cognitive Science Society.
Karl Pfleger, 1997. Categorical Boltzmann Machines. KSL Technical Report KSL-98-05, Computer Science Department, Stanford University.