Google tag (gtag.js)
Staff Scientist · OIST, Okinawa
Electronics Engineer turned AI researcher with a Ph.D. (with distinction) in Communications & Electronics from IPN Mexico. International career spanning Mexico, Italy, and Japan — now at the Okinawa Institute of Science and Technology working at the intersection of machine/deep learning and biomedical signals.
Research Impact
Screenshot taken August 2025. Visit the live profile for the latest metrics.
Focus Areas
Career
Research
01 / Transfer Learning
A transfer learning framework leveraging labeled PSG data to train models for awake/asleep classification and sleep staging on EEG from novel wearables. Addresses signal domain mismatch via latent space representations using deep learning, enabling accurate inference without manual annotations.
View details →
02 / Multimodal Biomarker
Integration of EEG and ECG signals from wearables to identify robust multimodal biomarkers for early disease detection. The framework fuses neurological and cardiovascular data, emphasizing real-time processing and artifact removal for continuous, non-invasive healthcare.
View details →
03 / Data Privacy
Bayesian Ridge Regression to tackle high-dimensional, correlated healthcare and smartwatch data under LDP. The "Castell" approach refines perturbation and aggregation, enabling accurate joint probability estimation with enhanced privacy safeguards.
View details →
04 / Anomaly Detection
Autoencoder-based models for real-time anomaly detection, tracking failures through elevated reconstruction errors. Identifies warning signs earlier and more accurately than supervised approaches, improving overall system reliability.
View details →
05 / Synthetic Data
A Bidirectional RNN with statistical stage to generate five types of synthetic biomedical signals (ECG, EEG, BCG, PPG, respiratory). Outperforms GAN-based techniques, addressing data scarcity crucial for medical diagnoses.
View details →06 / Forecasting
Region-based ARIMA model for COVID-19 spread prediction across six global regions with population data integration. Achieves an average RMSE of 144.81, with potential for improvement via climate and cultural variables.
View details →
07 / Computer Vision
Automatic face detection, segmentation, Gabor feature extraction, and PCA-reduced fuzzy clustering classification. Achieves ~97% accuracy with one region and up to 99% combining both, robust even under partial occlusion.
View details →Academic Output
IT = Invited Talk · J = Journal · B = Book · C = Conference. Full list: Full CV ↗
Transfer Learning from Clinical PSG to Real-World Wearables for Sleep Staging
19th BIOSTEC — HEALTHINF, pp. 548–555, ISBN 978-989-758-802-0
Opportunistic Teacher Forcing: Training RNNs on Wearable Signals with Inherent Missing Data
19th BIOSTEC — BIOSIGNALS, pp. 446–453, ISBN 978-989-758-802-0
Machine Learning-Driven Prediction of Heat Transfer Coefficients for Pure Refrigerants in Diverse Heat Exchanger Types
Journal of Experimental and Theoretical Analyses, 3(4), 32
Proceedings of the 24th International Conference on New Trends in Intelligent Software Methodologies, Tools and Techniques (SoMeT_25)
IOS Press 2025, Frontiers in AI and Applications 411
Differential Private Risk Factors Analysis of Polypharmacy
MDAI 2025, LNCS vol. 15957. Springer, Cham
Secure Aggregation of Smartwatch Health Data with LDP
SocialSec 2024, LNCS vol. 15565. Springer, Singapore
Comparative Analysis of Local Differential Privacy Schemes in Healthcare Datasets
Applied Sciences 2024, 14, 2864
Advances and Trends in Artificial Intelligence. Theory and Applications
Springer Singapore, 2024
Prediction of Heat Transfer Coefficient and Pressure Drop of Flow Boiling and Condensation Using Machine Learning
Journal of Physics: Conference Series, 2024
Meaningful Performance Analysis on Healthcare Data under Local Differential Privacy
SoMeT 2024: 398–411, ISBN 978-1-64368-538-0
Machine Learning-Based Approach to Correct Saturated Flow Boiling Heat Transfer Correlations
SoMeT 2024: 235–248, ISBN 978-1-64368-538-0
Ozone Responses to Reduced Precursor Emissions
Science of The Total Environment, 2023, 169180
An Efficient Local Differential Privacy Scheme Using Bayesian Ridge Regression
20th Annual Int. Conference on Privacy, Security & Trust (PST), Copenhagen, 2023
Webinar: Medical Big Data — Machine Learning
International Congress of Geoeconomic Engineering — Synthetic Biomedical Signals with ANNs
A Vulnerability in Video Anonymization — Privacy Disclosure from Face-Obfuscated Video
19th Annual Int. Conference on Privacy, Security & Trust (PST), Fredericton, 2022
Less Complexity One-Class Classification Approach Using Construction Error
Information Sciences, Vol. 560, pp. 217–234, 2021
A Novel Approach to Create Synthetic Biomedical Signals Using BiRNN
Information Sciences, Vol. 541, pp. 218–241, 2020
Recognition of ECG Signals Using Wavelet Based on Atomic Functions
Biocybernetics and Biomedical Engineering, Vol. 40, Issue 2, 2020
Forecasting of COVID-19 per Regions Using ARIMA Models and Polynomial Functions
Applied Soft Computing, Vol. 96, 2020
Heart Beat Recognition Using a Novel Preprocessing Scheme and Neural Networks
SoMeT 2020, Frontiers in AI and Applications, Vol. 327
Scheme Fuzzy Approach to Classify Skin Tonalities Through Geographic Distribution
J. Ambient Intelligence and Humanized Computing, 11, pp. 2859–2870, 2019
A Scheme to Classify Skin Through Geographic Distribution of Tonalities Using Fuzzy Based Classification Approach
SoMeT 2019, Frontiers in AI and Applications, Vol. 318
Facial Expression Recognition in Unconstrained Environment
SoMeT 2017, Frontiers in AI and Applications, Vol. 297, pp. 525–538
Facial Expression Recognition with Automatic Segmentation of Face Regions Using a Fuzzy Based Classification Approach
Knowledge-Based Systems, 2016, ISSN 0950-7051
A Facial Expression Recognition with Automatic Segmentation of Face Regions
SoMet 2015, Naples, Italy, LNCS pp. 529–540
A Supervised Classifier Scheme Based on Clustering Algorithms
CONCAPAN XXXIV, IEEE, Panama City, 2014