PaceRegression (original) (raw)

Class for building pace regression linear models and using them for prediction.

Under regularity conditions, pace regression is provably optimal when the number of coefficients tends to infinity. It consists of a group of estimators that are either overall optimal or optimal under certain conditions.

The current work of the pace regression theory, and therefore also this implementation, do not handle:

- missing values
- non-binary nominal attributes
- the case that n - k is small where n is the number of instances and k is the number of coefficients (the threshold used in this implmentation is 20)

For more information see:

Wang, Y (2000). A new approach to fitting linear models in high dimensional spaces. Hamilton, New Zealand.

Wang, Y., Witten, I. H.: Modeling for optimal probability prediction. In: Proceedings of the Nineteenth International Conference in Machine Learning, Sydney, Australia, 650-657, 2002.

BibTeX:

@phdthesis{Wang2000, address = {Hamilton, New Zealand}, author = {Wang, Y}, school = {Department of Computer Science, University of Waikato}, title = {A new approach to fitting linear models in high dimensional spaces}, year = {2000} }

@inproceedings{Wang2002, address = {Sydney, Australia}, author = {Wang, Y. and Witten, I. H.}, booktitle = {Proceedings of the Nineteenth International Conference in Machine Learning}, pages = {650-657}, title = {Modeling for optimal probability prediction}, year = {2002} }

Valid options are:

-D Produce debugging output. (default no debugging output)

-E The estimator can be one of the following: eb -- Empirical Bayes estimator for noraml mixture (default) nested -- Optimal nested model selector for normal mixture subset -- Optimal subset selector for normal mixture pace2 -- PACE2 for Chi-square mixture pace4 -- PACE4 for Chi-square mixture pace6 -- PACE6 for Chi-square mixture

ols -- Ordinary least squares estimator aic -- AIC estimator bic -- BIC estimator ric -- RIC estimator olsc -- Ordinary least squares subset selector with a threshold

-S Threshold value for the OLSC estimator