Search

Path Planning problems with Unmanned Aerial Vehicles (UAVs) are among the most studied knowledge areas in the related literature. However, few of them have been applied to groups of UAVs. The use of swarms allows to speed up the flight time and, thus, reducing the operational costs. When combined with Artificial Intelligence (AI) algorithms, a single system or operator can control all aircraft while optimal paths for each one can be computed. In order to introduce the current situation of these AI-based systems, a review of the most novel and relevant articles was carried out. This review was performed in two steps: first, a summary of the found articles; second, a quantitative analysis of the publications found based on different factors, such as the temporal evolution or the number of articles found based on different criteria. Therefore, this review provides not only a summary of the most recent work but it gives an overview of the trend in the use of AI algorithms in UAV swarms for Path Planning problems. The AI techniques of the articles found can be separated into four main groups based on their technique: reinforcement Learning techniques, Evolutive Computing techniques, Swarm Intelligence techniques, and, Graph Neural Networks. The final results show an increase in publications in recent years and that there is a change in the predominance of the most widely used techniques.

Facial biometrics play an essential role in the fields of law enforcement and forensic sciences. When comparing facial traits for human identification in photographs or videos, the analysis must account for several factors that impair the application of common identification techniques, such as illumination, pose, or expression. In particular, facial attributes can drastically change depending on the distance between the subject and the camera at the time of the picture. This effect is known as perspective distortion, which can severely affect the outcome of the comparative analysis. Hence, knowing the subject-to-camera distance of the original scene where the photograph was taken can help determine the degree of distortion, improve the accuracy of computer-aided recognition tools, and increase the reliability of human identification and further analyses. In this paper, we propose a deep learning approach to estimate the subject-to-camera distance of facial photographs: FacialSCDnet. Furthermore, we introduce a novel evaluation metric designed to guide the learning process, based on changes in facial distortion at different distances. To validate our proposal, we collected a novel dataset of facial photographs taken at several distances using both synthetic and real data. Our approach is fully automatic and can provide a numerical distance estimation for up to six meters, beyond which changes in facial distortion are not significant. The proposed method achieves an accurate estimation, with an average error below 6 cm of subject-to-camera distance for facial photographs in any frontal or lateral head pose, robust to facial hair, glasses, and partial occlusion., Departamento de Ciencias de la Computación y Sistemas Inteligentes

Financiado para publicación en acceso aberto: Universidade da Coruña/CISUG [Abstract] There has been steady growth in the adoption of Unmanned Aerial Vehicle (UAV) swarms by operators due to their time and cost benefits. However, this kind of system faces an important problem, which is the calculation of many optimal paths for each UAV. Solving this problem would allow control of many UAVs without human intervention while saving battery between recharges and performing several tasks simultaneously. The main aim is to develop a Reinforcement Learning based system capable of calculating the optimal flight path for a UAV swarm. This method stands out for its ability to learn through trial and error, allowing the model to adjust itself. The aim of these paths is to achieve full coverage of an overflight area for tasks such as field prospection, regardless of map size and the number of UAVs in the swarm. It is not necessary to establish targets or to have any previous knowledge other than the given map. Experiments have been conducted to determine whether it is optimal to establish a single control for all UAVs in the swarm or a control for each UAV. The results show that it is better to use one control for all UAVs because of the shorter flight time. In addition, the flight time is greatly affected by the size of the map. The results give starting points for future research, such as finding the optimal map size for each situation. Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. Funding for open access charge: Universidade da Coruña/CISUG. This work is supported by Instituto de Salud Carlos III, grant number PI17/01826 (Collaborative Project in Genomic Data Integration (CICLOGEN) funded by the Instituto de Salud Carlos III from the Spanish National plan for Scientific and Technical Research and Innovation 2013–2016 and the European Regional Development Funds (FEDER)—“A way to build Europe.”. This project was also supported by the General Directorate of Culture, Education and University Management of Xunta de Galicia ED431D 2017/16 and “Drug Discovery Galician Network” Ref. ED431G/01 and the “Galician Network for Colorectal Cancer Research” (Ref. ED431D 2017/23). This work was also funded by the grant for the consolidation and structuring of competitive research units (ED431C 2018/49) from the General Directorate of Culture, Education and University Management of Xunta de Galicia, and the CYTED network (PCI2018_093284) funded by the Spanish Ministry of Ministry of Innovation and Science. This project was also supported by the General Directorate of Culture, Education and University Management of Xunta de Galicia “PRACTICUM DIRECT” Ref. IN845D-2020/03 Xunta de Galicia; ED431D 2017/16 Xunta de Galicia; ED431G/01 Xunta de Galicia; ED431D 2017/23 Xunta de Galicia; ED431C 2018/49 Xunta de Galicia; IN845D-2020/03

Financiado para publicación en acceso aberto: Universidade da Coruña/CISUG [Abstract] Mussel farming is one of the most important aquaculture industries. The main risk to mussel farming is harmful algal blooms (HABs), which pose a risk to human consumption. In Galicia, the Spanish main producer of cultivated mussels, the opening and closing of the production areas is controlled by a monitoring program. In addition to the closures resulting from the presence of toxicity exceeding the legal threshold, in the absence of a confirmatory sampling and the existence of risk factors, precautionary closures may be applied. These decisions are made by experts without the support or formalisation of the experience on which they are based. Therefore, this work proposes a predictive model capable of supporting the application of precautionary closures. Achieving sensitivity, accuracy and kappa index values of 97.34%, 91.83% and 0.75 respectively, the kNN algorithm has provided the best results. This allows the creation of a system capable of helping in complex situations where forecast errors are more common. The authors want to acknowledge the support from INTECMAR, who have provide mostly data for this work and CESGA, who allows to conduct the tests on their installations. Funding for open access charge: Universidade da Coruña/CISUG. This work is supported by the “Collaborative Project in Genomic Data Integration (CICLOGEN) PI17/01826 funded by the Carlos III Health Institute in the context of the Spanish National Plan for Scientific and Technical Research and Innovation 2013–2016 and the European Regional Development Funds (FEDER)—”A way to build Europe.” This project was also supported by the General Directorate of Culture, Education and University Management of Xunta de Galicia (Ref. ED431G/01, ED431D 2017/16), the “Galician Network for Colorectal Cancer Research” (Ref. ED431D 2017/23), Competitive Reference Groups (Ref. ED431C 2018/49) and the Spanish Ministry of Economy and Competitiveness via funding of the unique installation BIOCAI (UNLC08-1E-002, UNLC13-13–3503) and the European Regional Development Funds (FEDER). Enrique Fernandez-Blanco would also like to thank NVidia corp., which granted a GPU used in this work for the preliminary tests Xunta de Galicia; ED431G/01 Xunta de Galicia; ED431D 2017/16 Xunta de Galicia; ED431D 2017/23 Xunta de Galicia; ED431C 2018/49

Colección: Contextos Universitarios Transformadores (CUT). Número 7 [Resumo] Unha das queixas máis frecuentemente realizadas polos estudantes ao longo dos anos sempre está referida coa avaliación dos traballos e prácticas. O método de corrección e puntuación percíbese habitualmente por parte destes como pouco claro ou inexistente ata a entrega. Por isto último, este traballo trata de explorar as diferentes alternativas dispoñibles dentro da ferramenta de teleformación da Universidade da Coruña, na procura de corrixir esta percepción. Para levalo a cabo, centrouse a atención en dúas materias do cuarto curso do Grao en Enxeñaría en Informática, que son cursadas polos estudantes en cuatrimestres consecutivos e polo que facilita a implantación dunha comparativa dentro dos mesmos grupos sen entorpecer o desenvolvemento normal das materias. A avaliación realizouse dende dous puntos de vista, a primeira, a percepción dos estudantes reflectida nas enquisas do curso dentro do programa de avaliación da calidade da universidade. Polo outro lado, o rendemento puramente académico sobre á cualificación académica media. Os resultados demostran entre os esquemas propostos para a cualificación, a rúbrica pública é a que produce os mellores resultados académicos ao longo do tempo, mentres que as guías de corrección son as mellores percibidas polos estudantes. [Abstract] Year after year, one recurrent complaint made by the students is always related to the grading of assignments and practical exercises. They usually perceive the correction and marking method as unclear or non-existent until the submission. For this last reason, this work explores the different alternatives available within the e-learning tool of the University of A Coruña, to correct this perception. To carry it out, we focused on two subjects in the fourth year of the Degree in Computer Science, which are taken by students in consecutive four-month terms, thereby, simplifying the implementation of a comparative study within the same groups without interfering with the normal development of the subjects. The evaluation was carried out from two points of view, firstly, the perception of the students reflected in the course surveys within the university's quality evaluation program. Secondly, the purely academic performance on the average of the academic qualifications. The results show that among the proposed qualification schemes, the public rubric is the one that produces the best academic results over time, while the correction guides are the best perceived by students. http://hdl.handle.net/2183/32493

Financiado para publicación en acceso aberto: Universidade da Coruña/CISUG [Abstract] Based on a solid mathematical background, this paper proposes a method for Symbolic Regression that enables the extraction of mathematical expressions from a dataset. Contrary to other approaches, such as Genetic Programming, the proposed method is deterministic and, consequently, does not require the creation of a population of initial solutions. Instead, a simple expression is grown until it fits the data. This method has been compared with four well-known Symbolic Regression techniques with a large number of datasets. As a result, on average, the proposed method returns better performance than the other techniques, with the advantage of returning mathematical expressions that can be easily used by different systems. Additionally, this method makes it possible to establish a threshold at the complexity of the expressions generated, i.e., the system can return mathematical expressions that are easily analyzed by the user, as opposed to other techniques that return very large expressions. This study is partially supported by Instituto de Salud Carlos III, grant number PI17/01826 (Collaborative Project in Genomic Data Integration (CICLOGEN) funded by the Instituto de Salud Carlos III from the Spanish National Plan for Scientific and Technical Research and Innovation 2013–2016 and the European Regional Development Funds (FEDER)—“A way to build Europe”. It was also partially supported by different grants and projects from the Xunta de Galicia [ED431D 2017/23; ED431D 2017/16; ED431G/01; ED431C 2018/49; IN845D-2020/03]. The authors thank the CyTED, Spain and each National Organism for Science and Technology for funding the IBEROBDIA project (P918PTE0409). In this regard, Spain specifically thanks the Ministry of Economy and Competitiveness for the financial support for this project through the State Program of I+D+I Oriented to the Challenges of Society 2017–2020 (International Joint Programming 2018), project (PCI2018-093284). Funding for open access charge: Universidade da Coruña/CISUG Xunta de Galicia; ED431D 2017/23 Xunta de Galicia; ED431D 2017/16 Xunta de Galicia; ED431G/01 Xunta de Galicia; ED431C 2018/49 Xunta de Galicia; IN845D-2020/03

Financiado para publicar en acceso aberto. Universidade da Coruña/CISUG [Abstract]: In recent years, mainly due to the application of smartphones in this area, research in human activity recognition (HAR) has shown a continuous and steady growth. Thanks to its wide range of sensors, its size, its ease of use, its low price and its applicability in many other fields, it is a highly attractive option for researchers. However, the vast majority of studies carried out so far focus on laboratory settings, outside of a real-life environment. In this work, unlike in other papers, progress was sought on the latter point. To do so, a dataset already published for this purpose was used. This dataset was collected using the sensors of the smartphones of different individuals in their daily life, with almost total freedom. To exploit these data, numerous experiments were carried out with various machine learning techniques and each of them with different hyperparameters. These experiments proved that, in this case, tree-based models, such as Random Forest, outperform the rest. The final result shows an enormous improvement in the accuracy of the best model found to date for this purpose, from 74.39% to 92.97%. Xunta de Galicia; ED431G 2019/01 Xunta de Galicia; ED481A 2020/003 Xunta de Galicia; ED431C 2022/46 Xunta de Galicia; ED431C 2018/49 Xunta de Galicia; ED431C 2021/53 This research was partially funded by MCIN/AEI/10.13039/ 501100011033, NextGenerationEU/PRTR, FLATCITY-POC , Spain [grant number P DC2021-121239-C31]; MCIN/AEI/10.13039/ 501100011033 MAGIST, Spain [grant number P ID2019-105221RB-C41]; Xunta de Galicia/FEDER-UE, Spain [grant numbers ED431G 2019/01 , ED481A 2020/003 , ED431C 2022/46 , ED431C 2018/49 and ED431C 2021/53 ]. Funding for open access charge: Universidade da Coruña/CISUG.

This article belongs to the Special Issue Applied Machine Learning in NIR Technology [Abstract] Among the bovine diseases, mastitis causes high economic losses in the dairy production system. Nowadays, detection under field conditions is mainly performed by the California Mastitis Test, which is considered the de facto standard. However, this method presents with problems of slowness and the expensiveness of the chemical-reactive process, which is deeply dependent on an expert’s trained eye and, consequently, is highly imprecise. The aim of this work is to propose a new method for bovine mastitis detection under field conditions. The proposed method uses a low-cost, smartphone-connected NIR spectrometer which solves the aforementioned problems of slowness, expert dependency and disposability of the chemical methods. This method uses spectra in combination with two k-Nearest Neighbors models. The first model is used to detect the presence of mastitis while the second model classifies the positive cases into weak and strong. The resulting method was validated by using a leave-one-out technique where the ground truth was obtained by the California Mastitis Test. The detection model achieved an accuracy of 92.4%, while the one classifying the severity showed an accuracy of 95%. This work is part of DINTA-UTMACH and RNASA-UDC research groups. This work is partially supported by Instituto de Salud Carlos III, grant number PI17/01826. It was also partially supported by different grants and projects from the Xunta de Galicia [ED431D 2017/23; ED431D 2017/16; ED431G/01; ED431C 2018/49; IN845D-2020/03]. Another source of support was the CYTED network (PCI2018\_093284) funded by the Spanish Ministry of Innovation and Science Xunta de Galicia; ED431D 2017/23 Xunta de Galicia; ED431D 2017/16 Xunta de Galicia; ED431G/01 Xunta de Galicia; ED431C 2018/49 Xunta de Galicia; IN845D-2020/03

Presented at the 4th XoveTIC Conference, A Coruña, Spain, 7–8 October 2021. [Abstract] The number of applications using unmanned aerial vehicles (UAVs) is increasing. The use of UAVs in swarms makes many operators see more advantages than the individual use of UAVs, thus reducing operational time and costs. The main objective of this work is to design a system that, using Reinforcement Learning (RL) and Artificial Neural Networks (ANNs) techniques, can obtain a good path for each UAV in the swarm and distribute the flight environment in such a way that the combination of the captured images is as simple as possible. To determine whether it is better to use a global ANN or multiple local ANNs, experiments have been done over the same map and with different numbers of UAVs at different altitudes. The results are measured based on the time taken to find a solution. The results show that the system works with any number of UAVs if the map is correctly partitioned. On the other hand, using local ANNs seems to be the option that can find solutions faster, ensuring better trajectories than using a single global network. There is no need to use additional map information other than the current state of the environment, like targets or distance maps. This research received no external funding.

Presented at the 4th XoveTIC Conference, A Coruña, Spain, 7–8 October 2021. [Abstract] Data processing and the use of machine learning techniques make it possible to solve a wide variety of problems. The great disadvantage of using this type of technology is the enormous amount of computation involved. This is why we have tried to develop an architecture that makes the best possible use of the resources available on each machine. The growth of cloud computing and the rise of virtualization techniques have led to a development that allows these tasks to be carried out in a more optimized way. The authors would like to thank the support from RNASA-IMEDIR group.

Presented at the 4th XoveTIC Conference, A Coruña, Spain, 7–8 October 2021. [Abstract] Knowing the chemical composition of a substance provides valuable information about it. That is why numerous techniques have been developed to try to obtain it. One of them is the Near Infrared Spectrometry technique, a non-destructive technique that analyzes the electromagnetic spectrum in search of waves of a certain length. The aim of this project is to combine this technology with machine learning techniques to try to detect the presence of milk, as well as the level of cocoa present in an ounce of chocolate. This has given satisfactory results in both cases, so it is considered that the combination of these techniques offers great possibilities. The authors would like to thank the support from RNASA-IMEDIR group.

[Abstract] Nowadays, among the Deep Learning works, there is a tendency to develop networks with millions of trainable parameters. However, this tendency has two main drawbacks: overfitting and resource consumption due to the low-quality features extracted by those networks. This paper presents a study focused on the scoring of sleeping EEG signals to measure if the increase of the pressure on the features due to a reduction of the number though different techniques results in a benefit. The work also studies the convenience of increasing the number of input signals in order to allow the network to extract better features. Additionally, it might be highlighted that the presented model achieves comparable results to the state-of-the-art with 1000 times less trainable and the presented model uses the whole dataset instead of the simplified versions in the published literature. This work has been partially funded by the Carlos III Health Institute and the European Regional Development Funds (FEDER) [PI17/01826]. It was also partially supported by different grants and projects from the Xunta de Galicia [ED431D 2017/23; ED431D 2017/16; ED431G/01; ED431C 2018/49] Xunta de Galicia; ED431D 2017/23 Xunta de Galicia; ED431D 2017/16 Xunta de Galicia; ED431G/01 Xunta de Galicia; ED431C 2018/49

[Abstract] Early detection is crucial to prevent the progression of Alzheimer’s disease (AD). Thus, specialists can begin preventive treatment as soon as possible. They demand fast and precise assessment in the diagnosis of AD in the earliest and hardest to detect stages. The main objective of this work is to develop a system that automatically detects the presence of the disease in sagittal magnetic resonance images (MRI), which are not generally used. Sagittal MRIs from ADNI and OASIS data sets were employed. Experiments were conducted using Transfer Learning (TL) techniques in order to achieve more accurate results. There are two main conclusions to be drawn from this work: first, the damages related to AD and its stages can be distinguished in sagittal MRI and, second, the results obtained using DL models with sagittal MRIs are similar to the state-of-the-art, which uses the horizontal-plane MRI. Although sagittal-plane MRIs are not commonly used, this work proved that they were, at least, as effective as MRI from other planes at identifying AD in early stages. This could pave the way for further research. Finally, one should bear in mind that in certain fields, obtaining the examples for a data set can be very expensive. This study proved that DL models could be built in these fields, whereas TL is an essential tool for completing the task with fewer examples. This work is supported by the “Collaborative Project in Genomic Data Integration (CICLOGEN)” PI17/01826 funded by the Carlos III Health Institute in the context of the Spanish National Plan for Scientific and Technical Research and Innovation 2013–2016 and the European Regional Development Funds (FEDER)—”A way to build Europe”. This project was also supported by the General Directorate of Culture, Education and University Management of Xunta de Galicia (Ref. ED431G/01, ED431D 2017/16), the “Galician Network for Colorectal Cancer Research” (Ref. ED431D 2017/23), Competitive Reference Groups (Ref. ED431C 2018/49) and the Spanish Ministry of Economy and Competitiveness via funding of the unique installation BIOCAI (UNLC08-1E-002, UNLC13-13-3503) and the European Regional Development Funds (FEDER) Xunta de Galicia; ED431G/01 Xunta de Galicia; ED431D 2017/16 Xunta de Galicia; ED431D 2017/23 Xunta de Galicia; ED431C 2018/49

This study seeks to support, through the use of Artificial Neural Networks (ANN), the decision to perform closings after days without sampling in the Vigo estuary. The opening and closing of the mussel production areas are based on the toxicity analysis of this bivalve&rsquo, s meat. Sometimes it is not possible to obtain the necessary data for effective closing. If there is evidence of an increase in toxicity levels, &ldquo, Precautionary Closings&rdquo, on mussel extraction is done. A small error in the forecast of the state of the areas could mean serious losses for the mussel industry and a huge risk for public health. Unlike in previous studies, this study aims to manage the state of the mussel production areas, whilst the others focused on predicting the harmful algae blooms. Having achieved test sensitivity values of 67.40% and test accuracy of 83.00%, these results may lead to new research that involves obtaining more accurate models that can be integrated into a support system.

Nowadays, consumers are more concerned about freshness and quality of food. Poultry egg storage time is a freshness and quality indicator in industrial and consumer applications, even though egg marking is not always required outside the European Union. Other authors have already published works using expensive laboratory equipment in order to determine the storage time and freshness in eggs. Oppositely, this paper presents a novel method based on low-cost devices for rapid and non-destructive prediction of egg storage time at room temperature ($23\pm1$��C). H&N brown flock with 49-week-old hens were used as source for the sampled eggs. Those samples were daily scanned with a low-cost smartphone-connected near infrared reflectance (NIR) spectrometer for a period of 22 days starting to run from the egg laid. The resulting dataset of 660 samples was randomly splitted according to a 10-fold cross validation in order to be used in a contrast and optimization process of two machine learning algorithms. During the optimization, several models were tested to develop a robust calibration model. The best model use a Savitzky Golay preprocessing technique, and an Artificial Neural Network with ten neurons in one hidden layer. Regressing the storage time of the eggs, tests achieved a coefficient of determination (R-squared) of $0.8319\pm0.0377$ and a root mean squared error (RMSE) of 1.97. Although further work is needed, this technique has shown industrial potential and consumer utility to determine the egg's freshness by using a low-cost spectrometer connected to a smartphone., 26 pages, 9 figures, 2 tables

[Abstract] The measurements of Near-Infrared (NIR) Spectroscopy, combined with data analysis techniques, are widely used for quality control in food production processes. This paper presents a methodology to optimize the calibration models of NIR spectra in four different stages in a sugar factory. The models were designed for quality monitoring, particularly °Brix and Sucrose, both common parameters in the sugar industry. A three stage optimization methodology, including pre-processing selection, feature selection and support vector machines regression metaparameters tuning, were applied to the spectral data divided by repeated cross-validation. Global models were optimized while endeavoring to ensure they are able to estimate both quality parameters with a single calibration, for the four steps of the process. The proposed models improve the prediction for the test set (unseen data) compared to previously published models, resulting in a more accurate quality assessment of the intermediate products of the process in the sugar industry.

[Abstract] It is a fact that, non-destructive measurement technologies have gain a lot of attention over the years. Among those technologies, NIR technology is the one which allows the analysis of electromagnetic spectrum looking for carbon-link interactions. This technology analyzes the electromagnetic spectrum in the band between 700 nm and 2500 nm, a band very close to the visible spectrum. Traditionally, the devices used to measure are utterly expensive and enormously bulky. That is why this project was focused on a portable spectrophotometer to make measures. This device is smaller and cheaper than the common spectrophotometer, although at the cost of a lower resolution. In this work, that device in combination with the use of machine learning was used to detect if a beer contains alcohol or it can be labeled as non-alcoholic drink. Xunta de Galicia; ED431G/01

[Abstract] Automatic detection of Alzheimer’s disease is a very active area of research. This is due to its usefulness in starting the protocol to stop the inevitable progression of this neurodegenerative disease. This paper proposes a system for the detection of the disease by means of Deep Learning techniques in magnetic resonance imaging (MRI). As a solution, a model of neuronal networks (ANN) and two sets of reference data for training are proposed. Finally, the goodness of this system is verified within the domain of the application.

Sleeping problems have become one of the major diseases all over the world. To tackle this issue, the basic tool used by specialists is the Polysomnogram, which is a collection of different signals recorded during sleep. After its recording, the specialists have to score the different signals according to one of the standard guidelines. This process is carried out manually, which can be highly time-consuming and very prone to annotation errors. Therefore, over the years, many approaches have been explored in an attempt to support the specialists in this task. In this paper, an approach based on convolutional neural networks is presented, where an in-depth comparison is performed in order to determine the convenience of using more than one signal simultaneously as input. Additionally, the models were also used as parts of an ensemble model to check whether any useful information can be extracted from signal processing a single signal at a time which the dual-signal model cannot identify. Tests have been performed by using a well-known dataset called expanded sleep-EDF, which is the most commonly used dataset as the benchmark for this problem. The tests were carried out with a leave-one-out cross-validation over the patients, which ensures that there is no possible contamination between training and testing. The resulting proposal is a network smaller than previously published ones, but which overcomes the results of any previous models on the same dataset. The best result shows an accuracy of 92.67\% and a Cohen's Kappa value over 0.84 compared to human experts., Comment: 20 pages, 4 figures, 4 tables

Traditionally, signal classification is a process in which previous knowledge of the signals is needed. Human experts decide which features are extracted from the signals, and used as inputs to the classification system. This requirement can make significant unknown information of the signal be missed by the experts and not be included in the features. This paper proposes a new method that automatically analyses the signals and extracts the features without any human participation. Therefore, there is no need for previous knowledge about the signals to be classified. The proposed method is based on Genetic Programming and, in order to test this method, it has been applied to a well-known EEG database related to epilepsy, a disease suffered by millions of people. As the results section shows, high accuracies in classification are obtained, Comment: Conference paper, doble-column, 7 pages, 1 figure, 3 tables

Signaling proteins are an important topic in drug development due to the increased importance of finding fast, accurate and cheap methods to evaluate new molecular targets involved in specific diseases. The complexity of the protein structure hinders the direct association of the signaling activity with the molecular structure. Therefore, the proposed solution involves the use of protein star graphs for the peptide sequence information encoding into specific topological indices calculated with S2SNet tool. The Quantitative Structure-Activity Relationship classification model obtained with Machine Learning techniques is able to predict new signaling peptides. The best classification model is the first signaling prediction model, which is based on eleven descriptors and it was obtained using the Support Vector Machines - Recursive Feature Elimination (SVM-RFE) technique with the Laplacian kernel (RFE-LAP) and an AUROC of 0.961. Testing a set of 3114 proteins of unknown function from the PDB database assessed the prediction performance of the model. Important signaling pathways are presented for three UniprotIDs (34 PDBs) with a signaling prediction greater than 98.0%., 19 pages, 6 figures, 3 tables

[Abstract] 1. The purpose of this work was to support decision-making in poultry farms by performing automatic early detection of anomalies in egg production. 2. Unprocessed data were collected from a commercial egg farm on a daily basis over 7 years. Records from a total of 24 flocks, each with approximately 20 000 laying hens, were studied. 3. Other similar works have required a prior feature extraction by a poultry expert, and this method is dependent on time and expert knowledge. 4. The present approach reduces the dependency on time and expert knowledge because of the automatic selection of relevant features and the use of artificial neural networks capable of cost-sensitive learning. 5. The optimum configuration of features and parameters in the proposed model was evaluated on unseen test data obtained by a repeated cross-validation technique. 6. The accuracy, sensitivity, specificity and positive predictive value are presented and discussed at 5 forecasting intervals. The accuracy of the proposed model was 0.9896 for the day before a problem occurs. Galicia. Consellería de Cultura, Educación e Ordenación Universitaria; GRC2014/049

[Abstract] This paper describes a new technique for signal classification by means of Genetic Programming (GP). The novelty of this technique is that no prior knowledge of the signals is needed to extract the features. Instead of it, GP is able to extract the most relevant features needed for classification. This technique has been applied for the solution of a well-known problem: the classification of EEG signals in epileptic and healthy patients. In this problem, signals obtained from EEG recordings must be correctly classified into their corresponding class. The aim is to show that the technique described here, with the automatic extraction of features, can return better results than the classical techniques based on manual extraction of features. For this purpose, a final comparison between the results obtained with this technique and other results found in the literature with the same database can be found. This comparison shows how this technique can improve the ones found. Instituto de Salud Carlos III; RD07/0067/0005 Xunta de Galicia; 10SIN105004PR

There is a need for a study of the complex diseases due to their important impact on our society. One of the solutions involves the theoretical methods which are fast and efficient tools that can lead to the discovery of new active drugs specially designed for these diseases. The Quantitative Structure - Activity Relationship models (QSAR) and the complex network theory become important solutions for screening and designing efficient pharmaceuticals by coding the chemical information of the molecules into molecular descriptors. This review presents the most recent studies on drug discovery and design using QSAR of several complex diseases in the fields of Neurology, Cardiology and Oncology.

[Resumen] El framework J2EE ha sido el gran dominador, durante mucho tiempo, en el desarrollo de aplicaciones empresariales. Esto hecho originó la aparición de un rico ecosistema de herramientas, manuales, tutoriales, etc., que explican las diferentes alternativas o peculiaridades a la hora de su implementación. La irrupción de .NET Framework, en el ámbito empresarial, ha producido una fuerte demanda de implementación de aplicaciones bajo dicha arquitectura. Sin embargo, la cantidad o calidad de la documentación disponible dista considerablemente con respecto a la existente, su principal alternativa (J2EE). Esta laguna de documentación, es especialmente visible y preocupante cuando se establece como objetivo dar a conocer los conceptos del framework, desde un punto de vista docente, a los futuros egresados del Grado en Ingeniería en Informática. Este trabajo describe el enfoque docente seguido, para alcanzar el citado objetivo de familiarizar a los alumnos con este framework alternativo y las prácticas habituales de modelo dentro de éste. Para ello, se basa principalmente de un conjunto de sencillos tutoriales con los que mostrar los fundamentos de la tecnología y dos aplicaciones completas (miniportal y minibank) en las que se muestra cómo aplicar patrones de diseño, a la hora de abordar una aplicación empresarial. [Abstract] For a long time, J2EE has been the dominating framework for the development of business applications. This fact resulted in a rich ecosystem of tools, manuals, tutorials, etc. that explain different implementation alternatives or peculiarities. The incursion of .NET Framework in the business environment has generated a strong demand of application implementation under this architecture. However, the quantity and quality of documents available significantly differs from its main alternative (J2EE). This documentation gap is especially visible and worrying when the objective is to teach the concepts of Framework, from a teacher’s point of view, to the future graduates of the Information Systems Engineering program. This paper describes the teaching approach used in order to achieve the goal of having the students become familiar with this alternative framework and the usual model practices within it. Thus, it is based mainly on a set of basic tutorials that show the foundations of technology and two complete applications (miniportal and minibank) explaining how to apply design patterns when developing a business solution. Keywords: systems engineering

[Abstract] Artificial Neuron–Glia Networks (ANGNs) are a novel bio-inspired machine learning approach. They extend classical Artificial Neural Networks (ANNs) by incorporating recent findings and suppositions about the way information is processed by neural and astrocytic networks in the most evolved living organisms. Although ANGNs are not a consolidated method, their performance against the traditional approach, i.e. without artificial astrocytes, was already demonstrated on classification problems. However, the corresponding learning algorithms developed so far strongly depends on a set of glial parameters which are manually tuned for each specific problem. As a consequence, previous experimental tests have to be done in order to determine an adequate set of values, making such manual parameter configuration time-consuming, error-prone, biased and problem dependent. Thus, in this paper, we propose a novel learning approach for ANGNs that fully automates the learning process, and gives the possibility of testing any kind of reasonable parameter configuration for each specific problem. This new learning algorithm, based on coevolutionary genetic algorithms, is able to properly learn all the ANGNs parameters. Its performance is tested on five classification problems achieving significantly better results than ANGN and competitive results with ANN approaches. Ministerio de Ciencia e Innovación; TIN2009-07707 Ministerio de Industria,Turismo y Comercio; TSI-020110-2009-53 Programa Iberoamericano de Ciencia y Tecnología para el Desarrollo; Red Ibero-NBIC 209RT0366 Galicia. Consellería de Economía e Industria; 10SIN105004PR Ministerio de Ciencia e Innovación; RD07/0067/005 Galicia. Consellería de Economía e Industria; 10MDS014CT Red Gallega de Investigación sobre Cáncer Colorrectal; Ref. 2009/58

[Abstract] Enzyme regulation proteins are very important due to their involvement in many biological processes that sustain life. The complexity of these proteins, the impossibility of identifying direct quantification molecular properties associated with the regulation of enzymatic activities, and their structural diversity creates the necessity for new theoretical methods that can predict the enzyme regulatory function of new proteins. The current work presents the first classification model that predicts protein enzyme regulators using the Markov mean properties. These protein descriptors encode the topological information of the amino acid into contact networks based on amino acid distances and physicochemical properties. MInD-Prot software calculated these molecular descriptors for 2415 protein chains (350 enzyme regulators) using five atom physicochemical properties (Mulliken electronegativity, Kang–Jhon polarizability, vdW area, atom contribution to P) and the protein 3D regions. The best classification models to predict enzyme regulators have been obtained with machine learning algorithms from Weka using 18 features. K* has been demonstrated to be the most accurate algorithm for this protein function classification. Wrapper Subset Evaluator and SVM-RFE approaches were used to perform a feature subset selection with the best results obtained from SVM-RFE. Classification performance employing all the available features can be reached using only the 8 most relevant features selected by SVM-RFE. Thus, the current work has demonstrated the possibility of predicting new molecular targets involved in enzyme regulation using fast theoretical algorithms. Galicia. Consellería de Economía e Industria, 10SIN105004PR Instituto de Salud Carlos III , PI13/00280

[Abstract] The recognition of seizures is very important for the diagnosis of patients with epilepsy. The seizure is a process of rhythmic discharge in brain and occurs rarely and unpredictably. This behavior generates a need of an automatic detection of seizures by using the signals of long-term electroencephalography (EEG) recordings. Due to the non-stationary character of EEG signals, the conventional methods of frequency analysis are not the best alternative to obtain good results in diagnostic purpose. The present work proposes a method of EEG signal analysis based on star graph topological indices (SGTIs) for the first time. The signal information, such as amplitude and time occurrence, is codified into invariant SGTIs which are the basis for the classification models that can discriminate the epileptic EEG records from the non-epileptic ones. The method with SGTIs and the simplest linear discriminant methods provide similar results to those previously published, which are based on the time-frequency analysis and artificial neural networks. Thus, this work proposes a simpler and faster alternative for automatic detection of seizures from the EEG recordings. Xunta de Galicia; 2007/127 Xunta de Galicia; 2007/144 Instituto de Salud Carlos III; PIO52048 Instituto de Salud Carlos III; RD07/0067/0005 Ministerio de Ciencia e Innovación; TIN2009—07707.

[Abstract] Aging and life quality is an important research topic nowadays in areas such as life sciences, chemistry, pharmacology, etc. People live longer, and, thus, they want to spend that extra time with a better quality of life. At this regard, there exists a tiny subset of molecules in nature, named antioxidant proteins that may influence the aging process. However, testing every single protein in order to identify its properties is quite expensive and inefficient. For this reason, this work proposes a model, in which the primary structure of the protein is represented using complex network graphs that can be used to reduce the number of proteins to be tested for antioxidant biological activity. The graph obtained as a representation will help us describe the complex system by using topological indices. More specifically, in this work, Randić’s Star Networks have been used as well as the associated indices, calculated with the S2SNet tool. In order to simulate the existing proportion of antioxidant proteins in nature, a dataset containing 1999 proteins, of which 324 are antioxidant proteins, was created. Using this data as input, Star Graph Topological Indices were calculated with the S2SNet tool. These indices were then used as input to several classification techniques. Among the techniques utilised, the Random Forest has shown the best performance, achieving a score of 94% correctly classified instances. Although the target class (antioxidant proteins) represents a tiny subset inside the dataset, the proposed model is able to achieve a percentage of 81.8% correctly classified instances for this class, with a precision of 81.3%. Galicia. Consellería de Economía e Industria; 10SIN105004PR Galicia. Consellería de Economía e Industria; O9SIN010105PR Ministerio de Economía y Competitividad; TIN-2009-07707

Signal classification is based on the extraction of several features that will be used as inputs of a classifier. The selection of these features is one of the most crucial parts, because they will design the search space, and, therefore, will determine the difficulty of the classification. Usually, these features are selected by using some prior knowledge about the signals, but there is no method that can determine that they are the most appropriate to solve the problem. This paper proposes a new technique for signal classification in which a Genetic Algorithm is used in order to automatically select the best feature set for signal classification, in combination with a kNN as classifier system. This method was used in a well known problem and its results improve those already published in other works.

EEG classification is a research topic that has attracted a lot of interest in recent years, as proven by the large number of papers published. To accomplish this task, a lot of classification systems such as Support Vector Machines (SVMs) or Artificial Neural Networks (ANNs) are used. However, Recurrent Artificial Neural Networks (RANNs) that allow using the previously computed results to generate the actual output have hardly been used, although intuitively they may seem to be very useful in this field. This article proposes the use of RANNs to solve a well-known problem: the detection of epileptic seizures in EEG signals. The results show that RANNs can work it out satisfactorily, with a higher accuracy than other techniques previously used.

The artificial embryogeny term overlaps all the models that try to adapt cellular properties into artificial models. This chapter presents a new model for artificial embryogeny that mimics the behaviour of biological cells, whose characteristics can be applied to solution of computational problems. The paper contains the theoretical development of the model and some test executed in an implementation of that model. The presented tests apply the model to simple structure generation and provide promising results with regard to its behaviour and applicability to more complex problems. The objective of the chapters is to be an introduction of the artificial embryogeny and shows an example of a model of these techniques.

This paper presents a model in the Artificial Embryogene (AE) framework. The presented system tries to model the main functions of the biological cell model. The main part of this paper describes the Gene Regulatory Network (GRN) model, which has a similar processing information capacity as Boole's Algebra. This paper also describes how to use it to perform the Iris Classification problem which is a pattern classification problem. The aim of this work is to show that the model can solve this kind of problems.

The main features of a new theoretical model inside the knowledge area called Artificial Embryogeny are described in this paper. Artificial Embryogeny is a term that identifies any model that uses embryological cells or embryological processes as inspiration. This chapter details the theoretical model and it also presents some its apllication to information processing problems. Specifically, this model was applied to solve classical problems such as pattern classification and pattern recognition problems. The Iris classification problem is the selected information processing problems presented in this paper. It must be remarked that a similar application was never been done with an artificial embryogeny model.

This paper presents the preliminary studies for the creation of a new tool to assist in medical diagnostic. The tool will help in the analysis of 2D-PAGE images. In order to create a 2D-PAGE image of an ideal patient—the patient could be healthy or ill—the tool will help us in the creation of an image that facilitates and speeds up future diagnostics. The creation of a master image has motivated the development of a tool to alignment gel images. The tool will make easier the correspondence among the proteins into the ideal image and the ones of a new image. Due to the fact that image registering process is quite complex, we use the Intel’s library OpenCV which provides functions to calculate optical flow and translation vectors. This library introduces into the project a set of variables unknown by the facultative. To solve this, an automatic selection of values for this set of variables is necessary. This last task is made with the Evolutionary Computation technique called Particle Swarm Optimization (Kennedy, R. & Eberhart, J. 1995)

This paper proposes a new evolutionary method for generating ANNs. In this method, a simple real-number string is used to codify both architecture and weights of the networks. Therefore, a simple GA can be used to evolve ANNs. One of the most interesting features of the technique presented here is that the networks obtained have been optimised, and they have a low number of neurons and connections. This technique has been applied to solve one of the most used benchmark problems, and results show that this technique can obtain better results than other automatic ANN development techniques.