Omega-Stat: An Environment for Implementing Intelligent Modeling Strategies
Omega-Stat, a new data analysis and modeling paradigm, is built on Lisp-Stat—an object-oriented statistical programming environment. It contains extensible, reusable-component libraries for performing data management, multivariate analyses, modeling, and dynamic graphics. A point-and-click user interface allows instant access to all objects, including analysis and graphics objects. Modeling is done by adding new model objects, i.e., extended datasets, to a tree structure originally containing prototypes for linear, generalized linear, and nonlinear models. Knowledge, and methods for accessing this knowledge, are embedded within model objects and edge objects linking these models. This representation allows the modeling process to be studied by following the analysis trails of expert analysts. The objective is to provide an expert consultant that is accessible as part of man/machine interaction. Modeling strategies can then be built into Omega-Stat, by using prior knowledge and data-analytic heuristics, to guide the process of constructing the model tree and the iterative search for an “optimal” model.
KeywordsModel Object Variable Object Dependency Tree Dynamic Graphic Graphical View
Unable to display preview. Download preview PDF.
- [CameronDixon92]Cameron, R. D. & Dixon A. H. (1992). Symbolic Computing with LISP. Englewood Cliffs, NJ: Prentice Hall.Google Scholar
- [ChambersHastie92]Chambers, J. M. & Hastie, T. J. (1992). Statistical Models in S. Pacific Grove, CA: Wadsworth.Google Scholar
- [Galfalvy94]Galvalvy, H.C. (1994). An object environment based on XLISP-STAT for multivariate analysis and dynamic graphics. Master’s Project Report, Department of Statistics and Computer Science, West Virginia University.Google Scholar
- [Hand93]Hand, D. J. (1993). Measurement scales as metadata. In D. J. Hand (Ed.), Artificial Intelligence Frontiers in Statistics. London: Chapman & Hall, 54–64.Google Scholar
- [Harner90]Harner, E. J. (1990). Interactively developing models based on the exponential family. In K. Berk and L. Malone (Eds.), Proceedings of the 21st Symposium on the Interface, 21,110–115.Google Scholar
- [Harner9l]Harner, E. J. (1991). An exploratory data analysis and modeling system based on the exponential family. Proceedings of the Statistical Computing Section of the Americal Statistical Association, 70–78.Google Scholar
- [Hasebe94]Hasebe, Y. (1994) A model parser and constructor for generalized linear models. Master’s Thesis, Department of Statistics and Computer Science, West Virginia University.Google Scholar
- [Michalewicz9l]Michalewicz, Z. (1991). Statistical and Scientific Databases. New York, NY: Ellis Horwood.Google Scholar
- [OldfordPeters86]Oldford, R. W. & Peters S. C. (1986). Implementation and study of statistical strategy. In W. A. Gale (Ed.), Aritficial Intelligence and Statistics. Reading, MA: Addison Wesley, 335–353.Google Scholar
- [Thisted86]Thisted, R. A. (1986). Representing statistical knowledge for expert data analysis systems. In W. A. Gale (Ed.), Aritficial Intelligence and Statistics. Reading, MA: Addison Wesley, 267–284.Google Scholar