A5046645490- Neural Networks and Applications
- EEG and Brain-Computer Interfaces
- Blind Source Separation Techniques
- Biometric Identification and Security
- Fuzzy Logic and Control Systems
- Time Series Analysis and Forecasting
- Digital Imaging for Blood Diseases
- Stock Market Forecasting Methods
- Advanced Chemical Sensor Technologies
- User Authentication and Security Systems
- Handwritten Text Recognition Techniques
- Image and Signal Denoising Methods
- Speech and Audio Processing
- Speech Recognition and Synthesis
- Image Retrieval and Classification Techniques
- Spectroscopy Techniques in Biomedical and Chemical Research
- Spectroscopy and Chemometric Analyses
- ECG Monitoring and Analysis
- Chaos control and synchronization
- Machine Fault Diagnosis Techniques
- Neural dynamics and brain function
- Machine Learning in Bioinformatics
- Gaze Tracking and Assistive Technology
- Neuroscience and Neural Engineering
- Medical Image Segmentation Techniques
National Institute of Astrophysics, Optics and Electronics
2015-2024
Hospital Universitario Virgen del Rocío
2024
Instituto de Biomedicina de Sevilla
2024
Universidad de Sevilla
2024
Biomedical Research Networking Center on Neurodegenerative Diseases
2024
Institute of Electrical and Electronics Engineers
2021
University of Catania
2021
Netherlands Institute for Radio Astronomy
2008
Universidad Popular Autónoma del Estado de Puebla
2007
Universidad de las Américas Puebla
2002-2006
This paper presents a novel approach on motor current signature analysis (MCSA) for broken bar fault detection of induction motors (IMs), using as input the signal measured from one three phases. Independent component (ICA) is used over Fourier-domain spectral signals obtained and its autocorrelation function. The standard deviation components within region interest (ROI) an ICA output was found to exhibit substantial differences between damaged healthy motors. Separation ROI in one, two,...
Broken bars detection on induction motors has been a topic of interest in recent years. Its is important due to the fact that failure silent and consequences it produces as power consumption increasing, vibration, introduction spurious frequencies electric line, among others, can be catastrophic. In this paper, use motor current signature analysis mathematical morphology detect broken under different mechanical load condition analyzed. The proposed algorithm first identifies then condition....
Electroencephalograms (EEG) have shown to be a useful approach measure the cognitive load in tasks where mental effort is involved. However, EEG signals present high variability among subjects as well non-stationary behavior, so that distributions samples of different are mismatched. Methods based on Unsupervised Domain Adaptation (UDA) been used an effective solution reduce such discrepancy, while ones leveraged by deep learning (D-UDA) improved classification results over shallow...
A brain computer interface (BCI) is a system that aims to control devices by analyzing signals patterns. In this work, convenient time-frequency representation (TFR) for visualizing ERD/ERS phenomenon (Event related synchronization and desynchronization) based on Hilbert transform spatial patterns addressed, wavelet feature extraction method motor imagery tasks presented. The vectors are constructed with four statistical energy parameters obtained from decomposition, the sub-band coding...
This paper describes an experiment on the detection of a P-300 rhythm from electroencephalographic signals for brain computer interfaces applications. The P300 evoked potential is obtained visual stimuli followed by motor response subject. EEG are with 14 electrodes Emotiv EPOC headset. Preprocessing includes denoising and blind source separation using Independent Component Analysis algorithm. detected through time-scale analysis based discrete wavelet transform (DWT). Comparison Short Time...
This paper presents a project on the development of cursor control emulating typical operations computer-mouse, using gyroscope and eye-blinking electromyographic signals which are obtained through commercial 16-electrode wireless headset, recently released by Emotiv. The position is controlled information from included in headset. clicks generated user's blinking with an adequate detection procedure based spectral-like technique called Empirical Mode Decomposition (EMD). EMD proposed as...
In this paper a novel neural network architecture for medium-term time series forecasting is presented. The proposed model, inspired on the Hybrid Complex Neural Network (HCNN) takes advantage of information obtained by wavelet decomposition and oscillatory abilities recurrent networks (RNN). prediction accuracy evaluated using 11 economic NN5 Forecasting Competition Artificial Networks Computational Intelligence, obtaining an average SMAPE 27%. model shows better mean performance in 56...
On a meta-learning process, the key is to build reliable meta-training data set, which requires best model for specific sample. In other hand, uncertainty of expected accuracy particular increases when depend on time. Then, during meta-learning, an accurate validation reliability involved models critical. This paper compares applicability two most used methods validating forecasting models: ten-fold and Monte Carlo cross validations. Experimental results, using time series NN3 tournament,...
Undiagnosed type 2 diabetes (T2D) remains a major public health concern. The global estimation of undiagnosed is about 46%, being this situation more critical in developing countries. Therefore, we proposed non-invasive method to quantify glycated hemoglobin (HbA1c) and glucose vivo. We developed technique based on Raman spectroscopy, RReliefF as feature selection method, regression feed-forward artificial neural networks (FFNN). spectra were obtained from the forearm, wrist, index finger 46...
In this paper we present preliminary results obtained from the application of morphological operator pecstrum, for extraction discriminating characteristics in leukocytes and similar artificial images. Experts have identified six categories leukocytes, very shape size, which makes them extremely difficult to distinguish automatically or even by non-expert humans. A feature vector based on a 7-component normalized area, nucleus - cytoplasm area ratio, was tested using 4 kinds recognizers:...
We propose the use of morphological pattern spectrum, or pecstrum, as base a biometric shape-based hand recognition system. The system receives an image right subject in unconstrained pose, which is captured with commercial flatbed scanner. According to pecstrum property invariance translation and rotation, does not require pegs for fixed position, simplifies acquisition process. This novel feature-extraction method tested using Euclidean distance classifier identification verification...
This paper presents a hand-shape biometric system based on novel feature extraction methodology using the morphological pattern spectrum or pecstrum.Identification experiments were carried out obtained vectors as an input to some recognition systems neural networks and support vector machine (SVM) techniques, obtaining in average identification of 98.5%.The verification case was analyzed through Euclidean distance classifier, acceptance rate (FAR) false rejection (FRR) for K-fold cross...
In this paper we report a novel application-basedmodel as suitable alternative for the classification and identification ofattacks on computer network, thus guarantee its safety from HTTP protocol-based malicious commands. The proposed model is built self-recurrentneural network based wavelets architecture with multidimensional radialwavelons, therefore suited to work online by analyzing non-linearpatterns in real time self-adjust changes input environment. Sixdifferent neural systems have...
Abstract Background Levodopa‐induced dyskinesia (LID) is a common adverse effect of levodopa, one the main therapeutics used to treat motor symptoms Parkinson's disease (PD). Previous evidence suggests connection between LID and disruption dopaminergic system as well genes implicated in PD, including GBA1 LRRK2 . Objectives Our goal was investigate effects genetic variants on risk time LID. Methods We performed genome‐wide association study (GWAS) analyses focused variants. also calculated...