A5019572093- Neural Networks and Applications
- Evolutionary Algorithms and Applications
- Metaheuristic Optimization Algorithms Research
- Rough Sets and Fuzzy Logic
- Advanced Multi-Objective Optimization Algorithms
- Machine Learning in Materials Science
- Time Series Analysis and Forecasting
- Computational Drug Discovery Methods
- Structural Health Monitoring Techniques
- Control Systems and Identification
- Fuzzy Logic and Control Systems
- Fault Detection and Control Systems
- Gene expression and cancer classification
- Protein Structure and Dynamics
- Data Mining Algorithms and Applications
- Music and Audio Processing
- Distributed and Parallel Computing Systems
- Machine Learning and ELM
- Image Retrieval and Classification Techniques
- Speech and Audio Processing
- Hydrological Forecasting Using AI
- Business Process Modeling and Analysis
- Meteorological Phenomena and Simulations
- Solar and Space Plasma Dynamics
- Voice and Speech Disorders
National Research Council Canada
2015-2024
National Academies of Sciences, Engineering, and Medicine
2006-2019
University of Puerto Rico-Mayaguez
2016
Mondragon University
2008
Universitat Politècnica de Catalunya
1997-2002
A long-standing challenge in the design of single atom alloys (SAAs), for catalytic applications, is determination a feature space that maximally correlates to molecular binding energies per Sabatier principle. The more representative underlying properties, greater predictive capability given machine learning (ML) algorithm. Moreover, diversity and range SAA impurities/sites examined, difficulty arriving at such feature. In this work, we undertake examine degree which adsorbate electronic...
The interest in automated analysis and classification of cough sounds has increased recent years, partly due to the worldwide COVID19 pandemic. To train such models, a large dataset is needed, however, it remains challenging find datasets that have expert-labelled diagnoses types. Data augmentation techniques been used machine learning models given limited data. Furthermore, ensures trained are invariant natural transformations data, measured from real environments/surroundings. This paper...
This paper presents an approach for constructing visual representations of high dimensional objective spaces using virtual reality. These arise from the solution multi-objective optimization problems with more than 3 functions which lead to Pareto fronts are difficult use. is preliminarily investigated both theoretically derived a test problem (DTLZ2) and practically obtained 4 knapsack via evolutionary algorithms like HLGA, NSGA, VEGA. The expected characteristics in terms relative sizes,...
This work reports on a comprehensive analysis of the predictive capacity and underlying physicochemical trends provided by d-band based electronic structure features as applied to single-atom alloys (SAAs). Taking CO adsorption energies at kink sites model framework, SAA are examined across range substrates with vastly differing intrinsic trends. Through this approach, it is demonstrated that properties can be highly transferable, often displaying atom-like behavior independent host...
Two medical data sets (Breast cancer and Colon cancer) are investigated within a visual mining paradigm through the unsupervised construction of virtual reality spaces using genetic programming classical optimization (for comparison purposes). The desired such that modified approach was proposed in order to generate programs representing vector functions. extension leads populations composed forests, instead single expression trees. No particular kind algorithm is required due generic nature...
Remote monitoring and measurement are valuable tools for medical applications they particularly important in the context of pandemic outbreaks, like current COVID-19. This paper presents an analysis sound measurements cough events from point view their predictive content with respect to identification different types cough, including positive COVID-19 cases. The data consisted a collection audio samples collected sources dry, wet, whooping coughs. Unsupervised supervised machine learning...
Abstract This paper (i) explores the internal structure of two quantum mechanics datasets (QM7b, QM9), composed several thousands organic molecules and described in terms electronic properties, (ii) further an inverse design approach to molecular consisting using machine learning methods approximate atomic composition molecules, QM9 data. Understanding characteristics this kind data is important when predicting from physical‐chemical properties designs. Intrinsic dimension analysis,...
Lighting on demand (LOD) is a high-resolution energy-efficient lighting control technique that dynamically adjusts the lights in space according to its occupancy. It controls individually such area receives higher light level determined by location of occupant(s) and, some cases, their tasks. In our LOD system, adjust individual above and around occupant track him or her while traversing circulation space. We developed three scenarios with different combinations levels occupancy background...
Multi-objective optimization is used for the computation of virtual reality spaces visual data mining and knowledge discovery. Two methods computing new are discussed: implicit explicit function representations. In first, images objects computed directly, in second, universal approximators (neural networks) obtained. The pros cons each approach discussed, as well their complementary character. NSGA-II algorithm requested to minimize two objectives: a similarity structure loss measure...