Implementation of evolutionary algorithms for deep architectures

Thumbnail Image
Other Title
Tirumala, Sreenivas Sremath
Author ORCID Profiles (clickable)
Conference Contribution - Paper in Published Proceedings
Ngā Upoko Tukutuku (Māori subject headings)
deep architectures
deep learning
evolutionary algorithms
ANZSRC Field of Research Code (2020)
Tirumala, S. S. (2014). Implementation of Evolutionary Algorithms for Deep Architectures. Antonio Lieto, Daniele P. Radicioni, Marco Cruciani:(Ed.), Proceedings of the Second International Workshop on Artificial Intelligence and Cognition (AIC 2014) (pp.164-171).
Deep learning is becoming an increasingly interesting and powerful machine learning method with successful applications in many domains, such as natural language processing, image recognition, and hand-written character recognition. Despite of its eminent success, limitations of traditional learning approach may still prevent deep learning from achieving a wide range of realistic learning tasks. Due to their flexibility and proven effectiveness, evolutionary learning techniques may therefore play a crucial role towards unleashing the full potential of deep learning in practice. Unfortunately, many researchers with a strong background on evolutionary computation are not fully aware of the state-ofthe-art research on deep learning. To meet this knowledge gap and to promote the research on evolutionary inspired deep learning techniques, this paper presents a comprehensive review of the latest deep architectures and surveys important evolutionary algorithms that can potentially be explored for training these deep architectures.
CEUR Workshop Proceedings
Link to ePress publication
Copyright holder
Copyright notice
Unless stated explicitely and in conformance to the legal Disclaimer of Sun SITE Central Europe (CEUR) and the legal Disclaimer of Technical University of Aachen (RWTH), the copyright for the workshop proceedings as a compilation, i.e., etc., is with the respective proceedings editors. The copyright for the individual items (subsuming any type of computer-represented files containing articles, software demos, videos, etc.) within a proceedings volume is owned by default by their respective authors. Copying of items, in particular papers, and proceedings volumes is permitted only for private and academic purposes. The permission for academic use implies an attribution obligation, i.e., you must properly cite the items that you use in your own published work. Modification of items is not permitted unless a suitable license is granted by its copyright owners. Copying or use for commercial purposes is forbidden unless an explicit permission is acquired from the copyright owners.
Copyright license
This item appears in: