 |
Data modeling here is defined as characterization of the input-output relationship between a (multivariate) data set and a response of interest. Typically, developing models of this relationship is complicated by the presence of spurious (nuisance) variables and noise as well as practical issues related to the time and effort of model development, validation, and deployment. Despite the difficulties, successful and timely model development can be very important in industrial settings for understanding, controlling, and exploiting relationships between available input variables and desired response behavior(s).
DataModeler is a soon-to-be-released package targeted at the efficient development of insight and models from multivariate data sets. The underlying techniques and algorithms have been applied for several years against industrial data modeling problems with recent developments focused on improving the ease-of-use, efficiency, and accessibility of the modeling process.
|
|
For the newest resources, visit Wolfram Repositories and Archives »
