Research: Discrete Multivariate Modeling

The methods used are also known in the systems literature as "reconstructability analysis" (RA). RA overlaps significantly with the fields of logic design and machine learning and with log-linear statistical modeling.The papers "Wholes and Parts in General Systems Methodology" and "An Overview of Reconstructability Analysis" listed below offer a concise review of RA methodology.

Projects

Theory/Methodology

OCCAM: RA software for data analysis & data mining

Occam3 (web accessible; try it out)

User manual (PDF)

EDA: Extended Dependency Analysis

Heuristic RA search for loopless models.
Download executable, sample files, and documentation (for Windows)

RA utility programs

Binning
Download Excel binning program

Time series masking
Download Unix mask program (tar file and read.me)

Klaus Krippendorff's information theory programs

Download zip file of Fortran code, sample files, and documentation

State-based modeling

Analysis based on events rather than variables

Preprocessing for neural nets and genetic algorithms

Variable selection & NN or GA structuring

Applications

Bio-medical data analysis

Decision Analysis

Pattern Recognition

Syntactic constraints of languages

Dynamics of cellular automata and random networks

Below is the lattice of structures for a 4-variable directed system with 1 dependent variable (output).
Boxes = relations; lines = variables;
bold lines = the dependent variable.

 

Papers

General

  • Zwick, M. (2004). "An Overview of Reconstructability Analysis." Kybernetes, vol. 33, No. 5/6, pp. 877-905. Abstract (HTML), (pdf)

  • Zwick, M. (2004). "Reconstructability Analysis with Fourier Transforms." Kybernetes, vol. 33, No. 5/6, pp. 1026-1040. Abstract (HTML), (pdf)

  • Grygiel, S., Zwick, M., and Perkowski, M. (2004). "Multi-Level Decomposition of Probabilistic Relations." Kybernetes, vol. 33, No. 5/6, pp. 948-961. Abstract (HTML), (pdf)

  • Willett, K., and Zwick, M. (2004). "A Software Architecture for Reconstructability Analysis." Kybernetes, vol. 33, No. 5/6, pp. 997-1008. Abstract (HTML), (pdf)

  • Zwick, M. (2001). "Wholes and Parts in General Systems Methodology." The Character Concept in Evolutionary Biology, Gunter Wagner, ed., Academic Press. pp. 237-256. (postscript / pdf)

  • Zwick, M. (1997). "Complexity and the Decomposability of Relations." Presented at the International Conference on Complex Systems, Nashua, New Hampshire. Abstract (HTML)

  • Anderson, D., and Zwick, M. (1997). "Resolution of Local Inconsistency in Identification." Presented at the International Institute for General Systems Studies, 2nd Workshop, Southwest Texas St. Univ., San Marcos, TX. Abstract (HTML)

  • Zwick, M. (1996). "Control Uniqueness In Reconstructability Analysis." Int. J. of General Systems, 24, pp.151-162. (postscript / pdf)

  • Zwick, M., and Zeitler, E. (1973). "Image Reconstruction from Projections." Optik, Vol. 38, No. 5, pp.550-565. (pdf)

Set-theoretic RA

  • Al-Rabadi, A., Zwick, M., and Perkowski, M. (2004). "A Comparison of Modified Reconstructability Analysis and Ashenhurst-Curtis Decomposition of Boolean Functions." Kybernetes, vol. 33, No. 5/6, pp. 933-947. Abstract (HTML), (pdf)

  • Al-Rabadi, A., and Zwick, M. (2004). "Reversible Modified Reconstructability Analysis of Boolean Circuits and Its Quantum Computation." Kybernetes, vol. 33, No. 5/6, pp. 921-932. Abstract (HTML), (pdf)

  • Al-Rabadi, A., and Zwick, M. (2004). "Modified Reconstructability Analysis for Many-Valued Functions and Relations." Kybernetes, vol. 33, No. 5/6, pp. 906-920. Abstract (pdf), (pdf)

  • Al-Rabadi, A. and Zwick M. (2004). "Enhancements to Crisp Possibilistic Reconstructability Analysis." International Journal of General Systems, vol. 33(4), pp. 361-382. Abstract (HTML), (pdf)

  • Zwick, M. (2001). "Wholes and Parts in General Systems Methodology." The Character Concept in Evolutionary Biology, Gunter Wagner, ed., Academic Press. pp. 237-256. (postscript / pdf)

  • Zwick, M., and Shu, H. (1995). "Set-Theoretic Reconstructability of Elementary Cellular Automata." Advances in Systems Science and Applications, 1, pp. 31-36. (postscript / pdf)

State-based RA

  • Zwick, M., and Johnson, M. S. (2004). "State-Based Reconstructability Analysis." Kybernetes, vol. 33, No. 5/6, pp. 1041-1052. Abstract (HTML), (pdf)

  • Johnson, M.S., and Zwick, M. (2000). "State-Based Reconstructability Modeling for Decision Analysis." Proceedings of The World Congress of the Systems Sciences and ISSS 2000, Allen, J .K. and Wilby, J.M. eds., Toronto, Canada: International Society for the Systems Sciences. Abstract (HTML), (postscript / pdf)

  • Zwick, M. (2001). "Wholes and Parts in General Systems Methodology." The Character Concept in Evolutionary Biology, Gunter Wagner, ed., Academic Press. (postscript / pdf)

  • Zwick, M. (1998). "Complexity Reduction in State-Based Modeling." International Conference on Complex Systems, Nashua, NH. Abstract (HTML)

Neural net and genetic algorithm preprocessing

  • Shervais, S., and Zwick, M. (2007). "Using Reconstructability Analysis for Input Variable Reduction: A Business Example." Proceedings of the 2007 IEEE International Conference on Information Reuse and Integration (IEEE IRI-2007), Las Vegas, July 13-15, 2007.
    (pdf)
  • Shervais, S., and Zwick, M. (2005). "Ordering Genetic Algorithm Genomes With Reconstructability Analysis: Discrete Models." Presented at IEEE Systems, Man, and Cybernetics meeting, Hawaii, Oct. 10-12. Abstract (HTML), (pdf)

  • Zwick, M., and Shervais, S. (2004). "Reconstructability Analysis Detection of Optimal Gene Order in Genetic Algorithms." Kybernetes, vol. 33, No. 5/6, pp. 1052-1062. Abstract (HTML), (pdf)

  • Shervais, S. and Zwick, M. (2003). "Ordering Genetic Algorithm Genomes With Reconstructability Analysis." International Journal of General Systems, Vol. 32(5), pp. 491-502. Abstract (HTML), (pdf)

  • Shervais, S. and Zwick, M. (2003). "Using Reconstructability Analysis to Select Input Variables for Artificial Neural Networks." International Joint Conference on Neural Nets, Portland OR, July 20-24. Abstract (HTML), (pdf)

  • Chambless, B., Lendaris, G., and Zwick, M. (2001). "An Information Theoretic Methodology for Prestructuring Neural Networks." International Joint Conference on Neural Networks (IJCNN), Washington DC. (pdf)

  • Chambless, B., and Scarborough, D. (2001). "Information-Theoretic Feature Selection for a Neural Behavioral Model." International Joint Conference on Neural Networks (IJCNN), Washington DC. (pdf)

  • Lendaris, G., Shannon, M., and Zwick, M. (1999). "Prestructuring Neural Networks for Pattern Recognition Using Extended Dependency Analysis." Invited paper, Applications and Science of Computational Intelligence II AeroSense '99, Orlando FL, SPIE. Entire paper (postscript / pdf)

  • Lendaris, G., Zwick M., and Mathia, K. (1993). "On Matching ANN Structure to Problem Domain Structure." Proceedings of World Congress on Neural Networks '93 (WCNN-93, Portland), Earlbaum/INNS. (pdf) (see also Prestructuring NNs based on structure in data)

Applications

  • Wilmot, B., Zwick, M., and McWeeney, S. (2008). "Reconstructability Analysis Detects Genetic Variation Associated with Gene Expression." Presented at the 12th QTL-MAS Workshop in Computational Genetics, Uppsala, Sweden, May 15-16. Abstract (pdf).

  • Wright, A., Ricciardi, T., and Zwick, M. (2005). "Application of Information-Theoretic Data Mining Techniques in a National Ambulatory Practice Outcomes Research Network." Presented at the American Medical Informatics Association annual symposium, Washington DC, Oct. 22-26. Abstract (HTML), (pdf)

  • Shervais, S., Zwick, M., and Kramer, P. (2005). "Reconstructability Analysis As A Tool For Identifying Gene-Gene Interactions In Studies Of Human Diseases." Presented at IEEE Systems, Man, and Cybernetics meeting, Hawaii, Oct. 10-12. Abstract (HTML), (pdf)

  • Shannon, T., and Zwick, M. (2004). "Directed Extended Dependency Analysis for Data Mining." Kybernetes, vol. 33, No. 5/6, pp. 973-983. Abstract (HTML), (pdf)

  • Chambless, B., and Scarborough, D. (2001). "Information-Theoretic Feature Selection for a Neural Behavioral Model." International Joint Conference on Neural Networks (IJCNN), Washington DC. (pdf)

  • Johnson, M.S., and Zwick, M., (2000). "State-Based Reconstructability Modeling for Decision Analysis." Proceedings of The WorldCongress of the Systems Sciences and ISSS 2000, Allen, J .K. and Wilby, J.M. eds., Toronto, Canada: International Society for the Systems Sciences. Abstract (HTML), (postscript / pdf)

  • Lendaris, G., Shannon, T., and Zwick, M. (1999). "Prestructuring Neural Networks for Pattern Recognition Using Extended Dependency Analysis." Invited paper, Applications and Science of Computational Intelligence II AeroSense '99, Orlando FL, SPIE. (postscript / pdf )

  • Zwick, M., and Shu, H., (1997). "Structure and Dynamics of Cellular Automata." International Institute for General Systems Studies, 2nd Workshop, Southwest Texas St. Univ., San Marcos, TX. Abstract (HTML)

  • Zwick, M., Shu, H., and Koch, R. (1995). "Information-Theoretic Mask Analysis of Rainfall Time-Series Data." Advances in Systems Science and Applications, 1, pp.154-159. (postscript / pdf)

  • Hosseini, J. C., Harmon, R. R., and Zwick, M. (1991). "An Information Theoretic Framework for Exploratory Multivariate Market Segmentation Research." Decision Sciences, Vol 22, pp 663-677. (pdf)

 

Relevant Courses

People

Faculty

Martin Zwick

Students

Pete Catching

Bjorn Chambless, ECE

Joe Fusion

Scott Mist

Tad Shannon

Ken Willett

Other Faculty
(collaborating or with related interests)

George Lendaris, SySc & ECE

Bart Massey, CS

James McNames, ECE

Tim Menzies, CS

Alan Mishchenko, ECE

Marek Perkowski, ECE

Nancy Perrin, SySc & Psych

Associated

Anas Al-Rabadi, ECE

Doug Anderson, Metro

Renato Carletti

Stan Grygiel, Ph.D. EE (PSU); Intel

Michael S. Johnson, Kaiser Permanente

Cliff Joslyn, Los Alamos National Labs

Steve Shervais, Ph.D. SySc (PSU)


Pages