Search: WFRF:(Kulcsár Balázs Adam 1975)
> (2020-2024) >
Controlled Descent ...
Controlled Descent Training
-
- Andersson, Viktor, 1995 (author)
- Chalmers tekniska högskola,Chalmers University of Technology
-
Varga, Balázs (author)
-
- Szolnoky, Vincent, 1995 (author)
- Chalmers tekniska högskola,Chalmers University of Technology
-
show more...
-
Syren, Andreas (author)
-
- Jörnsten, Rebecka, 1971 (author)
- Chalmers tekniska högskola,Chalmers University of Technology
-
- Kulcsár, Balázs Adam, 1975 (author)
- Chalmers tekniska högskola,Chalmers University of Technology
-
show less...
-
(creator_code:org_t)
- 2024
- 2024
- English.
-
In: International Journal of Robust and Nonlinear Control. - 1099-1239 .- 1049-8923. ; 34
- Related links:
-
https://research.cha... (primary) (free)
-
show more...
-
https://research.cha...
-
https://doi.org/10.1...
-
show less...
Abstract
Subject headings
Close
- In this work, a novel and model-based artificial neural network (ANN) training method is developed supported by optimal control theory. The method augments training labels in order to robustly guarantee training loss convergence and improve training convergence rate. Dynamic label augmentation is proposed within the framework of gradient descent training where the convergence of training loss is controlled. First, we capture the training behavior with the help of empirical Neural Tangent Kernels (NTK) and borrow tools from systems and control theory to analyze both the local and global training dynamics (e.g. stability, reachability). Second, we propose to dynamically alter the gradient descent training mechanism via fictitious labels as control inputs and an optimal state feedback policy. In this way, we enforce locally H2 optimal and convergent training behavior. The novel algorithm, Controlled Descent Training (CDT), guarantees local convergence. CDT unleashes new potentials in the analysis, interpretation, and design of ANN architectures. The applicability of the method is demonstrated on standard regression and classification problems.
Subject headings
- NATURVETENSKAP -- Matematik (hsv//swe)
- NATURAL SCIENCES -- Mathematics (hsv//eng)
- NATURVETENSKAP -- Data- och informationsvetenskap (hsv//swe)
- NATURAL SCIENCES -- Computer and Information Sciences (hsv//eng)
- TEKNIK OCH TEKNOLOGIER -- Elektroteknik och elektronik -- Reglerteknik (hsv//swe)
- ENGINEERING AND TECHNOLOGY -- Electrical Engineering, Electronic Engineering, Information Engineering -- Control Engineering (hsv//eng)
Keyword
- onvergent learning
- label augmentation
- Neural Tangent Kernel
- optimal labels
- label selection
- gradient decent training
Publication and Content Type
- art (subject category)
- ref (subject category)
Find in a library
To the university's database