Abstract
A two-step approach is introduced where a convolutional autoencoder is created that efficiently compresses the input data in the first step, and a convolutional neural network is created to classify the compressed data in the second step. The creation of networks in both steps is guided by by an evolutionary process, where new networks are constantly being generated by mutating members of a collection of existing networks. Additionally, a method is introduced that considers the trade-off between compression and information loss of different convolutional autoencoders. This is used to select the optimal convolutional autoencoder from among those evolved to compress the data for the second step.
The complete framework is implemented, tested on the popular CIFAR-10 data set, and the results are discussed. Finally, a number of possible directions for future work with this particular framework in mind are considered, including opportunities to improve its efficiency and its application in particular areas.
| Language | English |
|---|---|
| Title of host publication | Machine Learning, Optimization, and Data Science - 4th International Conference, LOD 2018, Revised Selected Papers |
| Editors | Giuseppe Nicosia, Giovanni Giuffrida, Giuseppe Nicosia, Panos Pardalos, Vincenzo Sciacca, Renato Umeton |
| Place of Publication | Cham |
| Publisher | Springer |
| Pages | 293-304 |
| Number of pages | 12 |
| ISBN (Electronic) | 978-3-030-13709-0 |
| ISBN (Print) | 978-3-030-13708-3 |
| DOIs | |
| State | Published - Feb 2019 |
| Event | 4th International Conference on Machine Learning, Optimization, and Data Science, LOD 2018 - Volterra, Italy Duration: 13 Sep 2018 → 16 Sep 2018 |
Publication series
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Volume | 11331 LNCS |
| ISSN (Print) | 0302-9743 |
| ISSN (Electronic) | 1611-3349 |
Conference
| Conference | 4th International Conference on Machine Learning, Optimization, and Data Science, LOD 2018 |
|---|---|
| Country | Italy |
| City | Volterra |
| Period | 13/09/18 → 16/09/18 |
Fingerprint
Keywords
- Convolutional autoencoders
- Convolutional neural networks
- Genetic algorithms
- Neuro-evolution
Cite this
}
Evolutionary construction of convolutional neural networks. / van Knippenberg, M.S.; Menkovski, V.; Consoli, Sergio.
Machine Learning, Optimization, and Data Science - 4th International Conference, LOD 2018, Revised Selected Papers. ed. / Giuseppe Nicosia; Giovanni Giuffrida; Giuseppe Nicosia; Panos Pardalos; Vincenzo Sciacca; Renato Umeton. Cham : Springer, 2019. p. 293-304 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11331 LNCS).Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review
TY - GEN
T1 - Evolutionary construction of convolutional neural networks
AU - van Knippenberg,M.S.
AU - Menkovski,V.
AU - Consoli,Sergio
PY - 2019/2
Y1 - 2019/2
N2 - Neuro-Evolution is a field of study that has recently gained significantly increased traction in the deep learning community. It combines deep neural networks and evolutionary algorithms to improve and/or automate the construction of neural networks. Recent Neuro-Evolution approaches have shown promising results, rivaling hand-crafted neural networks in terms of accuracy.A two-step approach is introduced where a convolutional autoencoder is created that efficiently compresses the input data in the first step, and a convolutional neural network is created to classify the compressed data in the second step. The creation of networks in both steps is guided by by an evolutionary process, where new networks are constantly being generated by mutating members of a collection of existing networks. Additionally, a method is introduced that considers the trade-off between compression and information loss of different convolutional autoencoders. This is used to select the optimal convolutional autoencoder from among those evolved to compress the data for the second step.The complete framework is implemented, tested on the popular CIFAR-10 data set, and the results are discussed. Finally, a number of possible directions for future work with this particular framework in mind are considered, including opportunities to improve its efficiency and its application in particular areas.
AB - Neuro-Evolution is a field of study that has recently gained significantly increased traction in the deep learning community. It combines deep neural networks and evolutionary algorithms to improve and/or automate the construction of neural networks. Recent Neuro-Evolution approaches have shown promising results, rivaling hand-crafted neural networks in terms of accuracy.A two-step approach is introduced where a convolutional autoencoder is created that efficiently compresses the input data in the first step, and a convolutional neural network is created to classify the compressed data in the second step. The creation of networks in both steps is guided by by an evolutionary process, where new networks are constantly being generated by mutating members of a collection of existing networks. Additionally, a method is introduced that considers the trade-off between compression and information loss of different convolutional autoencoders. This is used to select the optimal convolutional autoencoder from among those evolved to compress the data for the second step.The complete framework is implemented, tested on the popular CIFAR-10 data set, and the results are discussed. Finally, a number of possible directions for future work with this particular framework in mind are considered, including opportunities to improve its efficiency and its application in particular areas.
KW - Convolutional autoencoders
KW - Convolutional neural networks
KW - Genetic algorithms
KW - Neuro-evolution
UR - http://www.scopus.com/inward/record.url?scp=85063596775&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-13709-0_25
DO - 10.1007/978-3-030-13709-0_25
M3 - Conference contribution
SN - 978-3-030-13708-3
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 293
EP - 304
BT - Machine Learning, Optimization, and Data Science - 4th International Conference, LOD 2018, Revised Selected Papers
PB - Springer
CY - Cham
ER -