Call for Chapters:
Apologies for cross-posting
**********************************************************************************************************************
Book title: Large Scale Neuroevolution - On Swarm and Evolutionary Computing Approaches to Deep Learning
Publication: Studies in Computational Intelligence by Springer
Indexing to Web of Science, EI-Compendex, DBLP, SCOPUS, Google Scholar and Springerlink.
Aim and Scope: Deep Learning (DL) is becoming an increasingly interesting and powerful machine learning paradigm with successful applications in many areas. Currently, DL is considered as the best tool to extract the knowledge and insight from the huge volume of data. Deep learning architectures, most often employ feed-forward artificial neural networks with several hidden layers of neurons, a called Deep Neural Network (DNN). Despite their well-known benefits, DNNs present complex learning models whose architecture and parameterization are mostly hand-crafted. In addition, there is inefficiency in the training process of DNN due to the long training time for training. Moreover, there is another problem in finding the most accurate DNN in reasonable run-time due to the involvement of many parameters in the DNN model configurations and high dimension of the training datasets. On the other hand, due to the flexibility and proven effectiveness of evolutionary learning techniques and swarm intelligence, they may, therefore, play a crucial role towards unleashing the full potential of deep learning in practice and can optimize the machine learning models efficiently. However, many researchers with a strong background in evolutionary and swarm computation are not fully aware of the state-of-the-art research on deep learning. This book will review the significant role of meta-heuristic algorithms in optimizing DL for different aspects. In this book, the recent advances of applying meta-heuristics on DL will be reviewed and analyzed. In addition, some feasible research directions for bridging the gaps between meta-heuristics and DL will be presented. Also, this book aims to reflect the state-of-the-art while presenting a hybridization of the meta-heuristic algorithms and DL to evolve the parameters and the architecture of DNNs in order to maximize their classification accuracy, as well as maintaining a valid sequence of layers. Finally, challenges to be addressed and future directions of research will be identified. The objective of this book will be to address the new vision and motivation for the hybridization of meta-heuristic algorithms and DL and recent research topics and challenges in the hybridization scenario of meta-heuristic computing and DL for different applications.
List of topics: Focus of the chapters includes but is not limited to the following topics
· Evolutionary Computing for Deep Learning
· Swarm Intelligence Algorithms for Deep Learning
· Hybrid Evolutionary- Swarm Algorithms for Deep Learning
· Evolutionary Computing for hyper-parameter selection in deep learning
· Parameters Optimization of Deep Learning Models using Swarm Optimization
· Swarm optimization for hyper-parameter selection in deep neural networks
· Deep transfer leaning assisted with evolutionary computing
· Evolutionary fuzzy systems hybridization with deep neural systems
· Deep evolutionary reinforcement learning
· Multi and many-objective optimization approaches for deep learning
· Deep learning assisted design of evolutionary computing systems
· Application of deep evolutionary systems to computer vision, social networks, bioinformatics, and so on.
Volume Editors (Please feel free to contact by the e-mails provided, should you have any question):
Aboul Ella Hassanien, Cairo Univesrity, Egypt
Swagatam Das, Indian Statistical Institute, India
Francisco Hererra, University of Granada, Spain
**********************************************************************************************************************