Distributed Evolution of Deep Autoencoders

Jeff Hajewski; Suely Oliveira; Xiaoyu Xing

doi:10.1007/978-3-030-80119-9_6

Back

Book chapter

Distributed Evolution of Deep Autoencoders

Jeff Hajewski, Suely Oliveira and Xiaoyu Xing

Intelligent Computing, pp.133-153

Lecture Notes in Networks and Systems, Springer International Publishing

07/13/2021

DOI: 10.1007/978-3-030-80119-9_6

View Online

Abstract

Autoencoders have seen wide success in domains ranging from feature selection to information retrieval. Despite this success, designing an autoencoder for a given task remains a challenging undertaking due to the lack of firm intuition on how the backing neural network architectures of the encoder and decoder impact the overall performance of the autoencoder. In this work we present a distributed system that uses an efficient evolutionary algorithm to design a modular autoencoder. We demonstrate the effectiveness of this system on the tasks of manifold learning and image denoising. The system beats random search by nearly an order of magnitude on both tasks while achieving near linear horizontal scaling as additional worker nodes are added to the system.

Distributed deep learning

Evolutionary algorithms

Neural architecture search

Details

Title: Subtitle: Distributed Evolution of Deep Autoencoders
Creators: Jeff Hajewski
Suely Oliveira - University of Iowa
Xiaoyu Xing - Amazon (United States)
Resource Type: Book chapter
Publication Details: Intelligent Computing, pp.133-153
Series: Lecture Notes in Networks and Systems
DOI: 10.1007/978-3-030-80119-9_6
eISSN: 2367-3389
ISSN: 2367-3370
Publisher: Springer International Publishing; Cham
Language: English
Date published: 07/13/2021
Academic Unit: Computer Science; Mathematics
Record Identifier: 9984259469402771

Metrics

18 Record Views