TU Darmstadt / ULB / TUbiblio

Source Transformation for the Optimized Utilization of the Matlab Runtime System for Automatic Differentiation

Hück, Alexander and Bischof, Christian (2014):
Source Transformation for the Optimized Utilization of the Matlab Runtime System for Automatic Differentiation.
In: 3rd International Workshop on Computational Engineering, Stuttgart, Germany, October 6-10, 2014, [Online-Edition: http://ipvs.informatik.uni-stuttgart.de/SGS/ICCE2014/],
[Conference or Workshop Item]

Abstract

The ADiMat software is a tool that offers automatic differentiation of Matlab functions using a hybrid approach that combines source transformation and operator overloading. Performance tests showed a weakness with a derivative class of ADiMat. In a first step, the Matlab runtime environment is tested regarding function and method call overheads as well as property access overhead with Matlabs objects. An automatic method, transforming the XML based abstract syntax tree created with ADiMats toolchain through a set of stylesheets, was developed. The process completely removes the derivative object usage from any derivative function created with ADiMat. As a result, performance is improved considerable depending on the data container storing the derivative directions.

Item Type: Conference or Workshop Item
Erschienen: 2014
Creators: Hück, Alexander and Bischof, Christian
Title: Source Transformation for the Optimized Utilization of the Matlab Runtime System for Automatic Differentiation
Language: English
Abstract:

The ADiMat software is a tool that offers automatic differentiation of Matlab functions using a hybrid approach that combines source transformation and operator overloading. Performance tests showed a weakness with a derivative class of ADiMat. In a first step, the Matlab runtime environment is tested regarding function and method call overheads as well as property access overhead with Matlabs objects. An automatic method, transforming the XML based abstract syntax tree created with ADiMats toolchain through a set of stylesheets, was developed. The process completely removes the derivative object usage from any derivative function created with ADiMat. As a result, performance is improved considerable depending on the data container storing the derivative directions.

Divisions: 20 Department of Computer Science
20 Department of Computer Science > Scientific Computing
Event Title: 3rd International Workshop on Computational Engineering
Event Location: Stuttgart, Germany
Event Dates: October 6-10, 2014
Date Deposited: 14 Jul 2015 07:13
Official URL: http://ipvs.informatik.uni-stuttgart.de/SGS/ICCE2014/
Export:
Suche nach Titel in: TUfind oder in Google

Optionen (nur für Redakteure)

View Item View Item