NeCOL · An Innovative Methodology for Building Better Deep Learning Tools for Real Word Applications
Horizon 2020 — Marie Skłodowska-Curie Actions
- Duration
- 2018-09-10 → 2022-06-19
- EU contribution
- €239,191
- Participants
- 2
- Scheme
- MSCA-IF
Lines connect the coordinator with its partners.
Results in brief
NeCOL: An Innovative Methodology for Building Better Deep Learning Tools for Real Word Applications
Over the last years, the vast increase of digital data and the access to powerful computation resources have stimulated the fourth industrial revolution through the application of Artificial Intelligence (AI), specifically the application of deep neural networks (DNN), a.k.a. deep learning (DL). In spite of DL is being used to design breakthrough applications, the efficient design and training of DNNs is still an open problem and there is room for improvement. In this context, evolutionary computation (EC), which has been successfully used to solve hard-to-solve real-world optimization problems, emerges as an excellent tool to address DNN optimization. Here, we focus on recurrent neural networks (RNNs) and generative adversarial networks (GANs) Thus, this project has the relevant scientific goal of the development of a cutting edge DL methodology based on deep evolutionary neural networks (DENN), i.e., the combination of EC, specifically co-evolutionary algorithms (CEAs) and evolutionary algorithms (EAs), and DNN. Furthermore, we apply DENN to address applications of importance in present societies, such as cybersecurity and Smart Cities (SC). Hence the name of the project: Neural CO-evolutionary Learning or NeCOL. NeCOL will comprise a framework based on CEAs to deal with DL
Data: CORDIS, © European Union
Project objective
Nowadays, intelligent systems based on deep learning (DL) are latent in many aspects of our society. But the use of inadequate neural networks (NNs) architectures and the high computational costs required by DL limit its widespread use. Thus, advanced optimization methods (such as metaheuristics) may be applied to improve common DL methodologies, which in general use gradient based methods and apply complex engineering by hand.This project aims to define an efficient DL methodology, which is named Neural CO-evolutionary Learning (NeCOL), based on the marriage between co-evolutionary algorithms (CEAs) and recurrent NNs (RNNs). NeCOL will be used to automatically define RNNs of high (unseen) efficiency and efficacy, which will be adapted to explicit needs. It will be applied in two use cases of the highest value and relevance in EU: cybersecurity and Smart City. We focus on RNNs because they are applied to non-stationary data streams, as in our use cases.Despite EU efforts, China and the USA are the most productive countries in DL. Thus, EU must try harder to lead this compelling domain. This MSCA will support the candidate to master this new cutting-edge world-wide research, which will contribute to EU excellence and competitiveness. It will allow the candidate to get exceptional trainings from world class experts at the prestigious MIT that will be exploited at UMA and the priceless supervision of Prof. Alba (UMA) and Prof. O’Reilly (MIT).The applicant is the appropriate choice to successfully accomplish this research because he has a valuable expertise in modeling hard-to-solve real-world problems (as it is the case of RNNs optimization) and addressing them by using metaheuristics. The expected early high scientific impact of this research in the EU will open up the best possible career opportunities for him, preparing him to overwhelmingly compete for a solid permanent position at UMA and other possible destinations (even industry).
Original text from CORDIS.
Participants
- UNIVERSIDAD DE MALAGA · MALAGACoordinatorSpain
- MASSACHUSETTS INSTITUTE OF TECHNOLOGY · CambridgeUnited States
Links
Data: CORDIS, © European Union