

|
|
History
|
| |
|
Gepsoft Limited is a predictive modeling software company located in Bristol, United Kingdom.
Gepsoft was founded in 2000 to market the Gene Expression Programming (GEP)
technique invented by Dr. Candida Ferreira, founder and currently director of Gepsoft.
The first product to be released was a COM component (GEPSR 1.0 and 2.0) followed by
the desktop application APS (Automatic Problem Solver), which won the 2001 Microsoft
Portugal Software Award (Science). Gepsoft continues to develop this product, which
was renamed GeneXproTools (Gene Expression Programming Tools) after version 4.0.
The first version of APS, which only supported symbolic regression, was launched in February 2001.
Later that year Gepsoft released a second improved version that also introduced the Classification
algorithm. Version 3.0 was released in 2003 after a complete rewrite and introduction of new
algorithms for Time Series Prediction and the GEP-RNC algorithm for integrating random numerical
constants. Version 4.0 was released in July 2006 and supports for the first time unlimited
datasets (for both variables and samples) and new algorithms for Logic Synthesis and model simplification.
Two months later, in September of 2006, Gepsoft released the software GeneXproServer 1.0,
an add-on to GeneXproTools 4.0 and a powerful batch processor of GXPT runs that is
essential for the unattended processing of many different datasets.
|


|
|
Research
|
| |
|
Besides its main website (http://www.gepsoft.com/), Gepsoft maintains a dynamic research
website that tracks the developments in Gene Expression Programming:
http://www.gene-expression-programming.com/
The GEP algorithm outperforms the old Genetic Programming technique by a factor
of 100 - 60,000. This groundbreaking performance brings predictive
modeling to the average personal computer. These advances are due to a novel
structural organization that more closely mirrors the evolutionary mechanisms
in nature, making the design process not only thoroughly enjoyable but also highly efficient.
|
|
|
|
|
|
|
|