VarScreen

VarScreen is a free program which contains a variety of software tools useful for the developer of predictive models. These tools screen and evaluate candidates for predictors and targets. Some notable features of the program include:

Most operations involve just two quick steps: read the data and select the test to be performed. Program-supplied defaults are often satisfactory, and adjusting them is easy.
The program is fully multi-threaded, enabling it to take maximum advantage of modern multiple-core processors.
The most massively compute-intensive algorithms make use of any CUDA-enabled nVidia card in the user's computer. These widely available video cards turn an ordinary desktop computer into a super-computer, accelerating computations by several orders of magnitude.
Most tests print solo and unbiased p-values.

Available Tests

The Univariate Mutual Information test computes the mutual information between a specified target variable and each member of a specified set of predictor candidates.
The Bivariate Mutual Information test computes the mutual information between each of one or more specified target variables and each possible pair of predictors taken from a specified set of predictor candidates.
(New in Version 1.1) The Bivariate Mutual Information test now has the option of selecting candidates based on uncertainty reduction instead of mutual information.
(New in Version 1.2) A new test implements the algorithm of Peng, Long and Ding (2005) "Feature Selection Based on Mutual Information: Criteria of Max-Dependency, Max-Relevance, and Min Redundancy".
(New in Version 1.3) It develops hidden Markov models that fit all possible small subsets of a list of predictor candidates.
(New in Version 1.5) Stationarity in the mean is vital to most prediction schemes.
(New in Version 1.6) FREL (Feature Weighting as Regularized Energy-Based Learning) is a recent development that is useful in high-dimensionality, few-cases situations.
(New in Version 1.7) Principal components analysis via forward selection component analysis.
(New in Version 1.8) The Local Feature Selection algorithm identifies features whose predictive power is focused in relatively small areas of the feature domain.
(New in Version 1.9) Stepwise selection of predictive features enhanced in three important ways.
(New in Version 1.9) Nominal-to-ordinal conversion.
(New in Version 2.0) Print tables of autocorrelation and partial autocorrelation.
(New in Version 2.1) Improved test for break in mean of a time series.
(New in Version 2.2) Indicator selection for financial market prediction based on thresholded profit factor, with selection bias elimination.
(New in Version 2.3) Assorted useful plotting functions, and a bug fix for Version 2.2.
(New in Version 2.4, accelerated in Version 2.5) "Indicator selection by optimal profit factor" improvement with Monte-Carlo permutation test theory breakthrough.
(New in Version 3.1) The RANSAC algorithm trains a linear-quadratic prediction model by iteratively estimating the probability that each case is excessively dominated by noise.
(New in Version 3.3) Testing for mean reversion in a time series and the ability to directly read a market history file.
(New in Version 4.0) Three entropy tests for structure in a time series (Shannon, SVD K2).
(New in Version 4.1) Another entropy test, AP.
(New in Version 4.3) Several small bug fixes.
(New in Version 4.4) Generalized Shannon Entropy.

VarScreen is a work in progress. New screening algorithms will likely be added on a regular basis. Stay tuned.

Last but not least, please understand that VarScreen is an experimental program. It is provided free of charge to interested users for educational purposes only. In all likelihood this program contains errors and omissions. If you use this program for a purpose in which loss is possible, then you are fully responsible for any and all losses associated with use of this program. The developer of this program disclaims all responsibility for losses which the user may incur.

To download the manual only, click here.

For those who prefer a printed copy, you can purchase a paperback version of the online manual here. It's exactly the same as the pdf file that you can download from the link above.

To download the program (64-bit Windows) and manual (zipped), click here. You will need to change EXX to EXE after unzipping.