ExploreHD - High Dimensional Model Representations (HDMR)
ExploreHD implements the High Dimensional Model Representations (HDMR) technique pioneered by Prof. Herschel Rabitz and collaborators at Princeton University. HDMR explains the input-output behavior of a system using a hierarchical representation involving input variables acting individually and interacting in pairs as well as triples. The mathematical formulation of HDMR enables efficient global sensitivity analysis, readily providing information about how each input variable impacts each system output.
Whether the data comes from a computer model or experimental observation, ExploreHD helps you to understand and simplify the complex relationship between the inputs and outputs of your system. ExploreHD analyzes your data to determine the sensitivity of the system to the first, second and even third order interactions of the inputs, generating a Fully Equivalent Operational Model (FEOM) which captures the behavior of the system.
- Compatible with most data formats
- Includes both cut-HDMR and RS-HDMR algorithms
- Provides text and graphical views of analysis results
- Calculator utility allows prediction of system outputs at arbitrary inputs that may be specified individually or by importing a data file of several sets of inputs
- Post-processing capabilities include exporting data from ExploreHD to CSV file format
- Pre-processing capabilities include generating random distributions of inputs for creating input-output data sets from users systems (models or experiments).
- Mac OS X