Peter Lichodzijewski,
Ph. D.
Dalhousie University
Halifax, Nova Scotia
plichodz [at symbol] dal [period] ca
Research Interests
I'm interested in using Evolutionary Computation (EC) for model
building -- how to use concepts such as populations, genetic
variation, and natural selection to automatically build solutions to a
given problem. My work has dealt mainly with Genetic Programming (GP),
a branch of EC where the solutions take the form of (relatively) simple
programs (e.g., arithmetic expressions).
GP, and in general EC, can be used to find novel solutions to some
interesting and difficult problems (see the HUMIES page). Two domains
that I have experience with are classification and reinforcement
learning. Classification deals with the categorization of data, while
the goal of reinforcement learning is to learn useful policies through
interaction with an environment.
My current work looks at the issues of scalability and solution
transparency. A learning algorithm is viewed as scalable if it
can be successfully applied to very large problems (in terms of the
number of training instances or environmental states). Solution
transparency is important as it allows an end-user to gain
insight into the problem by analyzing the models that are developed. My
work investigates the usefulness of competitive coevolution in
overcoming the issue of scalability, while solution transparency is
addressed through bid-based problem decomposition. As it turns out, the
latter also provides an elegant way of building solutions that are
structured hierarchically which further supports the goal of
scalability (though perhaps, but not necessarily, at the expense of
solution transparency).
In early 2011, I defended my doctoral thesis "A
Symbiotic Bid-Based Framework for Problem Decomposition using Genetic
Programming" which delves into these ideas in detail. In
the near future I plan to release the associated source code, which for now
can obtained by contacting me by email.
Journal Publications
- Doucette, J., McIntyre, A., Lichodzijewski, P., Heywood,
M.I., "Symbiotic Coevolutionary Genetic Programming: A Benchmarking
Study
under Large Attribute Spaces," Genetic Programming and Evolvable
Machines, Vol. 13, No. 1, pp. 71-101, March 2012. (more)
- Lichodzijewski, P., Heywood, M.I., "Coevolutionary
Bid-Based Genetic
Programming for Problem Decomposition in Classification," Genetic
Programming and Evolvable Machines, Vol. 9, No. 4, pp. 331-365,
December 2008. (more)
- Curry, R., Lichodzijewski, P., Heywood, M.I., "Scaling
Genetic Programming to Large Datasets using Hierarchical Dynamic Subset
Selection," IEEE Transactions on Systems, Man, and Cybernetics - Part
B: Cybernetics, Vol. 37, No. 4, pp. 1065-1073, August 2007. (more)
Book Chapters
- Lichodzijewski, P., Heywood, M.I., "The Rubik's Cube and
GP Temporal Sequence Learning: An Initial Study," Genetic Programming
Theory and Practice, Vol 8., Ch. 3, pp. 35-54, 2011. (more)
- Doucette, J., Lichodzijewski, P., Heywood, M.I., "Evolving
Coevolutionary Classifiers under Large Attribute Spaces," Genetic
Programming Theory and Practice, Vol. 7, Ch. 3, pp. 37-54, 2010. (more)
Conference
Publications
- Doucette, J., Lichodzijewski, P., Heywood, M.I.,
"Hierarchical Task Decomposition through Symbiosis in Reinforcement
Learning," Proceedings of the 2012
Genetic and Evolutionary Computation Conference, July 2012. (to appear)
- Kelly, S., Lichodzijewski, P., Heywood, M.I., "On Meta
Actions and Hierarchical Symbiosis," Proceedings of the IEEE
Congress on Evolutionary Computation, 2012. (to appear)
- Lichodzijewski, P., Heywood, M.I., "Symbiosis,
Complexification and Simplicity under GP,"
Proceedings of the 2010
Genetic and Evolutionary Computation Conference, July 2010. (more)
- Heywood, M.I., Lichodzijewski, P., "Symbiogenesis as a
Mechanism for Building Complex Adaptive Systems: A Review," Proceedings
of the European Workshop on Evolutionary Algorithms and Complex
Systems, April 2010. (more)
- Lichodzijewski, P., Heywood, M.I., "Binary versus
Real-Valued Reward Functions under Coevolutionary Reinforcement
Learning," Proceedings of the 9th International Conference on
Artificial Evolution, October 2009.
- Alshammari, R., Lichodzijewski, P., Heywood, M.I.,
Zincir-Heywood, A. N., "Classifying SSH Encrypted Traffic with Minimum
Packet Header Features using Genetic Programming," Proceedings of the
2009 Genetic and Evolutionary Computation Conference, pp. 2539-2546,
July 2009. (more)
- Doucette, J., Lichodzijewski, P., Heywood, M.I.,
"Benchmarking Coevolutionary Teaming under Classification Problems with
Large Attribute Spaces," Proceedings of the
2009 Genetic and Evolutionary Computation Conference, pp. 1901-1902,
July 2009. (more)
- Lichodzijewski, P., Heywood, M.I., "Managing Team-Based
Problem Solving with Symbiotic Bid-Based Genetic Programming,"
Proceedings of the 2008
Genetic and Evolutionary Computation Conference, pp. 363-370, July
2008. (more)
- Lichodzijewski, P., Heywood, M.I., "Pareto-Coevolutionary
Genetic Programming for Problem Decomposition in Multi-Class
Classification," Proceedings of the 2007 Genetic and Evolutionary
Computation Conference, pp. 464-471, July 2007. (more)
- Lichodzijewski, P., Heywood, M.I., "GP Classifier Problem
Decomposition Using First-Price and Second-Price Auctions," Proceedings
of the 10th European Conference on Genetic Programming, pp. 137-147,
April 2007. (more)
- Allulli, L., Lichodzijewski, P., Zeh, N., "A Faster
Cache-Oblivious Shortest-Path Algorithm for Undirected Graphs with
Bounded Edge Lengths," Proceedings of the 18th ACM-SIAM Symposium on
Discrete Algorithms,
pp. 910-919, January 2007. (more)
- Wang, S., Lichodzijewski, P., "Boolean Genetic Programming
for Promoter Recognition in Eukaryotes," Proceedings of the IEEE
Congress on Evolutionary Computation, Vol. 1, pp. 683-690, September
2005. (more)
- Lichodzijewski, P., Heywood, M.I., Zincir-Heywood, A.N.,
"CasGP: Building Cascaded Hierarchical Models Using Niching,"
Proceedings of the IEEE Congress on Evolutionary Computation, Vol. 2,
pp.
1180-1187, September 2005. (more)
- Lichodzijewski, P., Heywood, M.I., Zincir-Heywood, A.N.,
"Cascaded GP Models for Data Mining," Proceedings of the IEEE Congress
on Evolutionary Computation, Vol. 2, pp. 2258-2264, July 2004. (more)
- Lichodzijewski, P., Wetmore L., Zincir-Heywood, A.N.,
"Pie: Scalable Routing in Peer-to-Peer Networks," Proceedings of the
IEEE Canadian Conference on Electrical and Computer Engineering, Vol.
1, pp.
51-54, May 2004. (more)
- Lichodzijewski, P., Zincir-Heywood, A.N., Heywood, M.I.,
"Dynamic Intrusion Detection Using Self-Organizing Maps," Proceedings
of the 14th Annual Canadian Information Technology Security Symposium,
May 2002.
- Lichodzijewski, P., Zincir-Heywood, A.N., Heywood, M.I.,
"Host-Based
Intrusion Detection Using Self-Organizing Feature Maps," Proceedings of
the IEEE International Joint Conference on Neural Networks, Vol. 2, pp.
1714-1719, May 2002. (more)
Theses
- Lichodzijewski, P., "A
Symbiotic
Bid-Based
Framework
for
Problem
Decomposition using Genetic
Programming," Ph.D. Thesis, Faculty of Computer Science, Dalhousie
University, February 2011. (more)
- Lichodzijewski, P., "Cascaded GP Models for Data Mining,"
Master's Thesis, Faculty of Computer Science, Dalhousie University,
August 2004.
- Lichodzijewski, P., "Network-Based Anomaly Detection Using
Self-Organizing Maps", Honours Thesis, Faculty of Computer Science,
Dalhousie University, April 2002.
|