Edgar Talavera Muñoz

Dr. Edgar Talavera Muñoz

Escuela Técnica Superior de Ingeniería de Sistemas Informáticos

Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 March 2023 in Neural Computing and Applications
Verified
Journal Article
Open access
PDF file
Guillermo IglesiasGuillermo Iglesias
Guillermo Iglesias
Edgar Talavera MuñozEdgar Talavera Muñoz
Edgar Talavera Muñoz
Ángel González-PrietoÁngel González-Prieto
Ángel González-Prieto
AMAlberto Mozo
Alberto Mozo
Sandra Gómez-CanavalSandra Gómez-Canaval
Sandra Gómez-Canaval
With the latest advances in deep learning-based generative models, it has not taken long to take advantage of their remarkable performance in the area of time series. Deep neural networks used to work with time series heavily depend on the size and consistency of the datasets used in training. These features are not usually abundant in the real world, where they are usually limited and often have constraints that must be guaranteed. Therefore, an effective way to increase the amount of data is by using data augmentation techniques, either by adding noise or permutations and by generating new synthetic data. This work systematically reviews the current state of the art in the area to provide an overview of all available algorithms and proposes a taxonomy of the most relevant research. The efficiency of the different variants will be evaluated as a central part of the process, as well as the different metrics to evaluate the performance and the main problems concerning each model will be analysed. The ultimate aim of this study is to provide a summary of the evolution and performance of areas that produce better results to guide future researchers in this field.

20 March 2023 in Computer Science Review
Verified
Journal Article
Open access
Guillermo IglesiasGuillermo Iglesias
Guillermo Iglesias
Edgar Talavera MuñozEdgar Talavera Muñoz
Edgar Talavera Muñoz
Alberto Díaz-ÁlvarezAlberto Díaz-Álvarez
Alberto Díaz-Álvarez

1 September 2015
Verified
Conference Paper
Closed access
JAJose J. Anaya
Jose J. Anaya
Edgar Talavera MuñozEdgar Talavera Muñoz
Edgar Talavera Muñoz
DGDavid Gimenez
David Gimenez
NBNURIA GOMEZ BLAS
NURIA GOMEZ BLAS
FJFelipe Jimenez
Felipe Jimenez
Jose NaranjoJose Naranjo
Jose Naranjo
Vehicle to vehicle (V2V) communications allow to sharing real-time information among the vehicles that circulate on the nearby road areas. These communications systems are today in an early stage of development and the effort are focused into two directions, on one hand, the development of the communications infrastructure to support the information exchange. On the other hand, the development of Advanced Driver Assistance Systems (ADAS) as well as safety systems that takes advantages of these communications. Additionally, the V2V communications systems are mainly focused to cars and trucks, showing less concern for Vulnerable Road Users, pedestrians, motorcyclists and cyclists. In this paper a novel ADAS is presented, focused to avoid accidents that involve motorcyclist and cyclists using V2V communications, incorporating them to the vehicular networking taking into account the intrinsic features of each collective and adapting to increase their safety.

1 July 2018 in Transportation Research Part F: Traffic Psychology and Behaviour
Verified
Journal Article
Closed access
Alberto Díaz-ÁlvarezAlberto Díaz-Álvarez
Alberto Díaz-Álvarez
Miguel ClavijoMiguel Clavijo
Miguel Clavijo
Felipe JimenezFelipe Jimenez
Felipe Jimenez
Edgar Talavera MuñozEdgar Talavera Muñoz
Edgar Talavera Muñoz
Francisco SerradillaFrancisco Serradilla
Francisco Serradilla

8 February 2022 in Scientific Reports
Verified
Journal Article
Open access
PDF file
AMAlberto Mozo
Alberto Mozo
ÁGÁngel González-Prieto
Ángel González-Prieto
APAntonio Pastor
Antonio Pastor
SGSandra Gómez-Canaval
Sandra Gómez-Canaval
Edgar Talavera MuñozEdgar Talavera Muñoz
Edgar Talavera Muñoz
Due to the growing rise of cyber attacks in the Internet, the demand of accurate intrusion detection systems (IDS) to prevent these vulnerabilities is increasing. To this aim, Machine Learning (ML) components have been proposed as an efficient and effective solution. However, its applicability scope is limited by two important issues: (i) the shortage of network traffic data datasets for attack analysis, and (ii) the data privacy constraints of the data to be used. To overcome these problems, Generative Adversarial Networks (GANs) have been proposed for synthetic flow-based network traffic generation. However, due to the ill-convergence of the GAN training, none of the existing solutions can generate high-quality fully synthetic data that can totally substitute real data in the training of ML components. In contrast, they mix real with synthetic data, which acts only as data augmentation components, leading to privacy breaches as real data is used. In sharp contrast, in this work we propose a novel and deterministic way to measure the quality of the synthetic data produced by a GAN both with respect to the real data and to its performance when used for ML tasks. As a by-product, we present a heuristic that uses these metrics for selecting the best performing generator during GAN training, leading to a novel stopping criterion, which can be applied even when different types of synthetic data are to be used in the same ML task. We demonstrate the adequacy of our proposal by generating synthetic cryptomining attacks and normal traffic flow-based data using an enhanced version of a Wasserstein GAN. The results evidence that the generated synthetic network traffic can completely replace real data when training a ML-based cryptomining detector, obtaining similar performance and avoiding privacy violations, since real data is not used in the training of the ML-based detector.

13 January 2018 in Energies
Verified
Journal Article
Open access
PDF file
Edgar Talavera MuñozEdgar Talavera Muñoz
Edgar Talavera Muñoz
Alberto Díaz-ÁlvarezAlberto Díaz-Álvarez
Alberto Díaz-Álvarez
Felipe JimenezFelipe Jimenez
Felipe Jimenez
Jose NaranjoJose Naranjo
Jose Naranjo
In recent years, vehicular communications systems have evolved and allowed for the improvement of adaptive cruise control (ACC) systems to make them cooperative (cooperative adaptive cruise control, CACC). Conventional ACC systems use sensors on the ego-vehicle, such as radar or computer vision, to generate their behavioral decisions. However, by having vehicle-to-X (V2X) onboard communications, the need to incorporate perception in the vehicle is drastically reduced. Thus, in this paper a CACC solution is proposed that only uses communications to make its decisions with the help of previous road mapping. At the same time, a method to develop these maps is presented, combining the information of a computer vision system to correct the positions obtained from the navigation system. In addition, the cut-in and cut-out maneuvers for a CACC platoon are taken into account, showing the tests of these situations in real environments with instrumented vehicles. To show the potential of the system in a larger-scale implementation, simulations of the behavior are provided under dense traffic conditions where the positive impact on the reduction of traffic congestion and fuel consumption is appreciated.