SETDiR: Scalable Extensible Toolkit for Dimensionality Reduction

Significant advances in high-throughput experiments and computational capability has ushered in an era of data-driven analysis and discovery. Recent years have seen the development of various statistical methods to reduce the noise, redundancy and dimensionality of this huge amount of data to make analysis more tractable. These range from traditional linear PCA based techniques to methods for non-linear model reduction like IsoMap. There is a need for a comprehensive software suite that can perform such analysis in a scalable, parallel way. This resulted in the development of SETDiR (Scalable Extensible Toolkit for Dimensionality Reduction).

SETDiR is a software tool developed to perform dimensionality reduction on a given high-dimensional set of data. The emphasis of the software is on various methods and techniques appropriate for a variety of high-dimensional data sets problems. These techniques can be used to efficiently unravel possible (non)-linear structures in the data. In addition, techniques to estimate the optimal dimensionality of the low-dimensional representation are included. The techniques are packaged into a modular, computational scalable software framework with a graphical user interface. This interface helps to separate out the mathematics and computational aspects from the applications, thus significantly enhancing utility to the scientific community.