Lehigh University CSE Dept. at Lehigh Resume ( pdf , ps , text ) Send E-Mail to me Teaching Statement( pdf , ps , text ) Research Statement( pdf , ps , text ) Unix/X Linux C/C++/OPP Perl Emacs Tech. Reps and Bibs. Artificial Intelligence Machine Learning Neural Networks Data Mining Integrating Connectionism with Symbolism & Rule Integration and Extraction in NNs __o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__ Integrating Symbolism with Connectionism The topic of my Ph.D. thesis was : "Converting a Trained Neural Network to a Decision Tree Decext. - Decision Tree Extractor". I introduced new splitting techniques, a discretizatin and a pruning technique for extracting classical (C4.5 like) decision tree from trained feedforward neural networks (I used backpropagation networks).I called the new algorithm Another contribution of my thesis is a new Feature Subset Selection technique. Real world classification applications usually have many features. This increases the complexity of the classification task. Choosing a subset of the features will increase accuracy and reduce complexity of the acquired knowledge. The new technique has been tested on real world and artificial datasets and the results were compared with existing methods. It was shown that the new method chooses good subsets by searching fewer states than the existing methods. Here is a link to tarred gzipped source code with the executables for RedHat 7.1: FSS-src.tar.gz <./FSS-src.tar.gz>. I am currently working on a paper on FSS. I will put the draft here as soon as it is ready. I will also try to cleanup the source code and make it leaner and more usable. Let me know if you use the code and have any problems or comments. II will also add some documentation to the program. Click on these links to access my dissertation (in pdf , postscript formats). Here are my presentatin slides (in pdf , postscript formats). Here is another presentation I had in Prof. Pottenger 's Data Mining Class at Lehigh (in power point format) When I first started working on my thesis I intented to work on "Rule Integration and Extraction in Neural Networks". THen I decided to skip the Rule Integration part. The reason is if you try to integrate rules into Neural Networks you have to create specialized enural networks which make sth enet less effective. My Ph.D. Advisor was Prof. Donald J. Hillman . I have submitted my Ph.D. proposal <./integrated/proposal.html> (gzipped postscript copy <./integrated/proposal.ps.gz>) in the fall semester of 1995. You can also check my proposal slides <./integrated/slides.html> and gzipped postscript copy <./integrated/slides.ps.gz> of the slides. I am also working on a review paper on "Integration of symbolism with connectionism, rule integration and extraction in neural networks." <./integrated/paper.html>. You can get the gzipped postscript copy <./integrated/paper.ps.gz> of the paper too. Bibliographies and Papers :-) Bibliography O List of papers (some with links) related to integration of symbolism with connectionism, rule integration and extration in neural networks, connectionist neural networks, expert networks etc. (Gzipped postscript copy <./integrated/references.ps.gz>). :-) Dissertations O List of dissertations (some with links)related to integration of symbolism with connectionism, rule integration and extration in neural networks, connectionist neural networks, expert networks etc. :-) bibliography on the Domain/Theory Combination literature :-) Publications of the UW ML research Group O This page contains relevant information about publications by the members of the Machine Learning Research Group (MLRG) at the University of Wisconsin - Madison. Publications include Journal Articles, Conference Papers, Book Chapters, Workshop Papers Tech Reports, PhD Theses, ASCII file containing abstracts . :-) Evolutionary Design of Neural Architectures - A Preliminary Taxonomy and Guide to Literature Balakrishnan, K. and Honavar, V. (Postscript File) or the bibiography (.bib File) :-) Publications of AI Group in Computer SCience Department at Univ. of Geneva. :-) Ron Sun's Hybrid Systems Bibliography :-) Selected publications by Lorien Pratt :-) Selected Publications on neural network methods by Lorien Pratt :-) List of papers on Novel Neural Network Techniques developed in the ESPRIT-project NEUFODI :-) Towell's papers at Siemens FTP site :-) Prof. Henry Tirri's FTP directory of the papers - University of Helsinki :-) Publications and technical reports of DSI Neural Network Group :-) sighi directory O This directory contains technical reports as a public service to the hybrid intellegence and advanced technology community. Researchers may place electronic versions of their preprints in this directory, announce availability, and other interested researchers can rapidly retrieve and print the postscripts. :-) Hybrid Neural Network and Expert Systems Bibliography by Larry Medsker :-) KRUST publications - KRUST is an automated knowledge refinement system for rule-based systems. :-) Theory refinement papers of the Machine Learning Research Group at The University of Texas at Austin. :-) Prof. Lacher's recent publications :-) Papers on transfer and learning to learn :-) Dr. Sun's bibliography of connectionist models with symbolic processing. :-) Prof. Joachim Diederich's Publications (Date 6/7/95) __o0o__ Technical Reports FTP directory :-) Robert Andrew's Selected Publications :-) Ismail Taha's publications :-) Hybrid Intelligent Systems Bibliography __o0o__ Hybrid Intelligent Systems Books :-) Selected publications of COMPLEX SYSTEMS COMPUTATION GROUP (COSCO) - University of Helsinki, Finland - Prof. Tirri's group :-) Steve Galant's publications :-) (book) EXPLANATION-BASED NEURAL NETWORK LEARNING: A LIFELONG LEARNING APPROACH :-) Papers from the Workshop on Integrating Multiple Learned Models for Improving and Scaling Machine Learning Algorithms :-) List and abstracts of Papers presente in ECAI '96 Workshop on Neural Networks and Structured Knowledge (NNSK) :-) Dr. Sethi's recent publications on Neural Networks. :-) Special Issue: Knowledge-Based Neural Networks {Knowledge-Based Systems, 8(6), December 1995} Guest Editor: LiMin Fu, University of Florida (Gainesville, USA) o Introduction to knowledge-based neural networks L Fu o Dynamically adding symbolically meaningful nodes to knowledge-based neural networks D W Opitz and J W Shavlik o Recurrent neural networks and prior knowledge for sequence processing: A constrained nondeterministic approach P Frasconi, M Gori, and G Soda o Initialization of neural networks by means of decision trees I Ivanova and M Kubat o Extension of the temporal synchrony approach to dynamic variable binding in a connectionist inference system N S Park, D Robertson, and K Stenning o Hybrid modeling in pattern recognition and control Jim Bezdek o Survey and critique of techniques for extracting rules from trained artificial neural networks R. Andrewsm J Diederich, and A B Tickle __o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__ Researchers :-) Robert Andrews O Lecturer at the School of Information Systems in the Faculty of Information Technology in Brisbane, Australia :-) Kevin J. Cherkauer O Ph.D. Student, Department of Computer Sciences, University of Wisconsin - Madison Advisor : Jude Shavlik Research Interests : Machine learning, input representations, decision trees, genetic algorithms, neural networks, artificial intelligence, parallel computing :-) Mark Craven O Ph.D. Student, Department of Computer Sciences, University of Wisconsin - Madison Advisor : Jude Shavlik Research Interests : + Machine learning + Extracting symbolic representations from non-symbolic classifiers + Issues of comprehensibility + Computational learning theory + Hypothesis boosting + Constructive induction + Computational biology + Recognizing signals in DNA + Recognizing protein-coding regions in DNA + Protein classification :-) Joachim ("Joe") Diederich O Professor of Neurocomputing - Computing Science, QUT Research Interests : + Rule-Extraction from Artificial Neural Networks + Hybrid Artificial Intelligence/Neural Network Systems + Computational Learning Theory + Biologically-Plausible Neural Networks + Neural Networks and Parallel Computation :-) Thomas G. Dietterich O Professor in Computer Science Department, Oregon State University. Research Interests: + Skill acquisition and speedup learning methods, including explanation-based learning and reinforcement learning with applications to space shuttle scheduling and cognitive modeling; + Learning in the presence of prior knowledge applied to automatic calibration of ecosystem models. + Supervised learning for problems with large numbers of classes; + Learning from ambiguous examples, with applications to rational drug design; :-) C. Lee Giles's Home Page O Senior Research Scientist, Computer Sciences, NEC Research Institute, & Adjunct Associate Professor, Institute for Advanced Computer Studies, U of Maryland, College Park, MD Research Interests: + Dynamically-driven recurrent neural networks + Hybrid systems + Neural network and AI applications in communications, computing and computers, multi-media, adaptive control, system identification, networks, language processing and time series. :-) Prof. Vasant Honavar O Assistant Professor, AI Research Group , Computer Science Department, Iowa State University. Research Interests : + Artificial Intelligence + Artificial Neural Networks + Knowledge-Based Systems + Machine Learning + Neural Architectures for Knowledge Representation and Inference + And more ..... :-) Prof. Lacher's Home Page O Professor and chair in Computer Science at Florida State University. :-) Prof. Zoran Obradovic O Assistant Professor, Electrical Engineering and Computer Science , Washington State University . :-) Rich Maclin O Assistant Professor, Department of Computer Science, University of Minnesota-Duluth Advisor : Jude Shavlik Research Interests : Machine learning, neural networks, reinforcement learning, learning by being told, protein secondary-structure prediction :-) Christian W. Omlin O Visiting Scientist, Nec Research Institute. Research Interests : + Machine learning + Symbolic learning + Neural networks + Genetic algorithms. :-) David Opitz O Assistant Professor, Department of Computer Science, University of Minnesota-Duluth Advisor : Jude Shavlik Research Interests : Artificial intelligence, machine learning, neural networks, genetic algorithms, theory refinement, ensemble methods, anytime learning :-) Lori Pratt O Assistant Professor, Dept. of Mathematical and Computer Sciences, Center for Robotics and Intelligent Systems,Colorado School of Mines. Research Interests : + Transfer between neural networks + Applied neural networks + Neural networks + AI in telecommunications + Machine Learning + Artificial Intelligence :-) Prof. Jude W. Shavlik's Home Page O Associate Professor, Computer Sciences Department, University of Wisconsin Research Interests : Machine learning systems that combine the strengths of symbolic approaches to artificial intelligence with those of connectionist AI. :-) Dr. Ron Sun O Assistant Professor, University of Alabama Computer Science Department Some of Dr. Sun's Papers Research Interests : + Artificial Intelligence + Cognitive Science + Reasoning, Learning, and Connectionist Models + Integrating traditional symbolic AI and the new connectionist AI, + Exploring symbolic processing capabilities in connectionist models :-) Ismail A. Taha O Ph.D candidate at The University of Texas at Austin in the Dept. of Electrical and Computer Engineering. Research Interests : + Hybrid Intelligent Architectures that combine the paradigms of Connectionist models, Statistics, and Knowledeg Based Systems. :-) Prof. Henry Tirri O Prof. in Department of Computer Science, University of Helsinki. Research Interests : + Bayesian Learning + Neural computing + Evolutionary computing + Transaction processing + Representation transformations __o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__ Software :-) NRC - AUSIM NN-ES integrator FTP directory O AUSIM is a simulator for neural networks and massively parallel AI. :-) Neither - NEITHER is a propositional theory refinement system that will modify a incomplete or incorrect rule base so as to make it consistent with a set of input training examples. __o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__ Research Group and Projects :-) MIX - Modular Integration of Connectionist and Symbolic Processing in Knowledge Based Systems - ESPRITIII Basic Research Project - CUI Switzarland :-) MIX Project page at Loria France :-) HAREM - A Hybrid Architecture for Risk Evaluation and Management at CUI :-) The KRUST Home Page an automated knowledge refinement system for rule-based systems __o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__ Resources :-) Rule Extraction From Artificial Neural Networks Mailing List. o To post an article to the list send mail to annrules@fit.qut.edu.au o To subscribe send mail to annrules-request@fit.qut.edu.au (Subject: subscribe) o To remove your name from the list simply send mail to annrules-request@fit.qut.edu.au (Subject: unsubscribe) o To receive each day's messages bundled into a single posting send mail to annrules-digest-request@fit.qut.edu.au (Subject: subscribe) :-) Hybrid Models - Learning and Architectures Mailing List. o To subscribe, send email to listproc@cs.ua.edu with message body (not subject) subscribe hybrid-list your_name and your_last_name After subscription you can post to the list by sending e-mail to hybrid-list@cs.ua.edu . o To unsubscribe, send email to listproc@cs.ua.edu with message body (not subject) unsubscribe hybrid-list o To receive digests (accumulated messages) instead of as they arrive: send email to listproc@cs.ua.edu with no subject but a one-line message SET hybrid-list MAIL DIGEST Hybrid list archives :-) Evolutionary Design of Neural Architectures Mailing List. To subscribe, send email to majordomo@cs.iastate.edu with message body (not subject) "subscribe gann-list". :-) Intelligent Hybrid Systems - University College London - maintaned by Shukhdev Khebbal and Suran Goonatilake - editors of the book "Intelligent Hybrid Systems" :-) Hybrid Intelligent Systems Home Page O Created and Maintained by Sumit Virmani and Larry Medsker at American University. :-) Cognitive Systems: Integrated and Hybrid Architectures and Algorithms Page - maintained by Vasant Honavar. Conferences, Workshops, Tutorials, Schools and Call for Papers :-) NIPS*96 POST-CONFERENCE WORKSHOP - RULE-EXTRACTION FROM TRAINED NEURAL NETWORKS Snowmass (Aspen), Colorado, USA Fri December 6th, 1996 :-) ECAI '96 Workshop on Neural Networks and Structured Knowledge (NNSK) August 12, 1996 in Budapest, Hungary :-) Workshop on Integrating Multiple Learned Models for Improving and Scaling Machine Learning Algorithms to be held in conjunction with AAAI-96 (collocated with KDD-96, UAI-96, and IAAI-96) Portland, Oregon August 4-5th, 1996 :-) Integrating Multiple Learned Models for Improving and Scaling Machine Learning Algorithms held in conjunction with AAAI-96 (collocated with KDD-96, UAI-96, and IAAI-96) Portland, Oregon August 4-5th, 1996 :-) ECAI '96 Workshop on Neural Networks and Structured Knowledge (NNSK) to be held on August 12, 1996 during the 12th European Conference on Artificial Intelligence from August 12-16, 1996 in Budapest, Hungary :-) AUTUMN SCHOOL on HYBRID SYMBOLIC CONNECTIONIST SYSTEMS O Ravensburg -- Weingarten (Germany), September 25--28, 1996 organized by ESPRIT BASIC RESEARCH PROJECT 9119 (MIX) "Modular Integration of Connectionist and Symbolic Processing in Knowledge-Based Systems" :-) Call for papers O Special Issue of the Machine Learning Journal on Inductive Transfer, Lorien Pratt and Sebastian Thrun, Guest Editors :-) RULE-EXTRACTION FROM TRAINED NEURAL NETWORKS O AISB-96 WORKSHOP Society for the Study of Artificial Intelligence and Simulation of Behaviour (SSAISB) April 2, 1996 :-) Post-NIPS*95 Workshop, December 1-2, 1995, Vail, Colorado O "Learning to Learn: Knowledge Consolidation and Transfer in Inductive Systems" :-) Workshop at KI-96: "Knowledge Representation in Neural Networks" Sept.17 - Sept.19, 1996 in Dresden, Germany __o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__o0o__ [ E-MAIL ] To contact me, send e-mail to olcay@ptd.net . ------------------------------------------------------------------------ Olcay Boz , olcay@ptd.net Last modified: Mon Nov 2 02:52:31 EST 1998