Deep Reinforcement Learning for Optimization at Early Design Stages

Lorenzo Servadei; Jin Hwa Lee; Jose A.Arjona Medina; Michael Werner; Sepp Hochreiter; Wolfgang Ecker; Robert Wille

doi:10.1109/MDAT.2022.3145344

Deep Reinforcement Learning for Optimization at Early Design Stages

Lorenzo Servadei, Jin Hwa Lee, Jose A.Arjona Medina, Michael Werner, Sepp Hochreiter, Wolfgang Ecker, Robert Wille

Computation, Information and Technology

Research output: Contribution to journal › Article › peer-review

3 Scopus citations

Abstract

Deep reinforcement learning is shown to improve the design cost of hardware char63software interfaces within an industrial design framework. Based on optimization preferences specified by a designer, the proposed approach generates optimized solutions.

Original language	English
Pages (from-to)	43-51
Number of pages	9
Journal	IEEE Design and Test
Volume	40
Issue number	1
DOIs	https://doi.org/10.1109/MDAT.2022.3145344
State	Published - 1 Feb 2023

Keywords

Combinatorial Optimization
Design Automation
Early Design Stages
Machine Learning
Reinforcement Learning

Access to Document

10.1109/MDAT.2022.3145344

Cite this

@article{ae1d551e22ed47b1865c2a614be86efe,

title = "Deep Reinforcement Learning for Optimization at Early Design Stages",

abstract = "Deep reinforcement learning is shown to improve the design cost of hardware char63software interfaces within an industrial design framework. Based on optimization preferences specified by a designer, the proposed approach generates optimized solutions.",

keywords = "Combinatorial Optimization, Design Automation, Early Design Stages, Machine Learning, Reinforcement Learning",

author = "Lorenzo Servadei and Lee, \{Jin Hwa\} and Medina, \{Jose A.Arjona\} and Michael Werner and Sepp Hochreiter and Wolfgang Ecker and Robert Wille",

note = "Publisher Copyright: {\textcopyright} 2013 IEEE.",

year = "2023",

month = feb,

day = "1",

doi = "10.1109/MDAT.2022.3145344",

language = "English",

volume = "40",

pages = "43--51",

journal = "IEEE Design and Test",

issn = "2168-2356",

publisher = "IEEE Computer Society",

number = "1",

}

TY - JOUR

T1 - Deep Reinforcement Learning for Optimization at Early Design Stages

AU - Servadei, Lorenzo

AU - Lee, Jin Hwa

AU - Medina, Jose A.Arjona

AU - Werner, Michael

AU - Hochreiter, Sepp

AU - Ecker, Wolfgang

AU - Wille, Robert

PY - 2023/2/1

Y1 - 2023/2/1

N2 - Deep reinforcement learning is shown to improve the design cost of hardware char63software interfaces within an industrial design framework. Based on optimization preferences specified by a designer, the proposed approach generates optimized solutions.

AB - Deep reinforcement learning is shown to improve the design cost of hardware char63software interfaces within an industrial design framework. Based on optimization preferences specified by a designer, the proposed approach generates optimized solutions.

KW - Combinatorial Optimization

KW - Design Automation

KW - Early Design Stages

KW - Machine Learning

KW - Reinforcement Learning

UR - https://www.scopus.com/pages/publications/85123688050

U2 - 10.1109/MDAT.2022.3145344

DO - 10.1109/MDAT.2022.3145344

M3 - Article

AN - SCOPUS:85123688050

SN - 2168-2356

VL - 40

SP - 43

EP - 51

JO - IEEE Design and Test

JF - IEEE Design and Test

IS - 1

ER -

Deep Reinforcement Learning for Optimization at Early Design Stages

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this