A5077595408- Advanced Multi-Objective Optimization Algorithms
- Hydrological Forecasting Using AI
- Advanced Data Storage Technologies
- Meteorological Phenomena and Simulations
- Atmospheric and Environmental Gas Dynamics
- Machine Learning and Data Classification
- Advanced Combustion Engine Technologies
- Metalloenzymes and iron-sulfur proteins
- Hydrology and Watershed Management Studies
- Biodiesel Production and Applications
- Fault Detection and Control Systems
- Computational Physics and Python Applications
- Advanced Control Systems Optimization
- Genomics and Rare Diseases
- Metaheuristic Optimization Algorithms Research
- Advanced Optimization Algorithms Research
- Machine Learning and Algorithms
- Radiomics and Machine Learning in Medical Imaging
- Reservoir Engineering and Simulation Methods
- Advanced Neural Network Applications
- Oceanographic and Atmospheric Processes
- Model Reduction and Neural Networks
- Folate and B Vitamins Research
- Advanced Manufacturing and Logistics Optimization
- Mitochondrial Function and Pathology
National Renewable Energy Laboratory
2024-2025
Great Ormond Street Hospital
2023-2024
University College London
2023-2024
Lawrence Berkeley National Laboratory
2018-2022
Cytoplasmic and nuclear iron-sulfur (Fe-S) enzymes that are essential for genome maintenance replication depend on the cytoplasmic Fe-S assembly (CIA) machinery cluster acquisition. The core of CIA consists a complex CIAO1, MMS19 FAM96B. physiological consequences loss function in components pathway have thus far remained uncharacterized. Our study revealed patients with biallelic CIAO1 developed proximal axial muscle weakness, fluctuating creatine kinase elevation, respiratory...
MATSuMoTo is the MATLAB Surrogate Model Toolbox for computationally expensive, black-box, global optimization problems that may have continuous, mixed-integer, or pure integer variables. Due to black-box nature of objective function, derivatives are not available. Hence, surrogate models used as cheap approximations expensive function in order guide search improved solutions. computational expense doing a single evaluation, goal find optimal solutions within very few evaluations. The...
Today there has never been a more profound codependence and synergetic convergence, such as the one between energy IT sectors, industries will need to work together closely meet today's growing data center power demands.
Mitochondrial diseases frequently affect the brain leading to severe and disabling neurological symptoms. The heteroplasmic m.3243A>G mutation in MT-TL1, encoding mt-tRNALeu, is responsible for ~80% of mitochondrial encephalomyopathy, lactic acidosis, stroke-like episodes (MELAS), which one most characteristic syndromes, disability early death. There are no animal models harbouring this provide precise mechanistic insights informing therapeutic interventions. Here, we generated a human...
Cytoplasmic and nuclear iron-sulfur enzymes that are essential for genome maintenance replication depend on the cytoplasmic assembly (CIA) machinery cluster acquisition. Here we report patients with biallelic loss of function in CIAO1 , a key CIA component, develop proximal axial muscle weakness, fluctuating creatine kinase elevation respiratory insufficiency. In addition, they present CNS symptoms including learning difficulties neurobehavioral comorbidities, along iron deposition deep...
Abstract The Generative Adversarial Network (GAN) is a powerful and flexible tool that can generate high-fidelity synthesized data by learning. It has seen many applications in simulating events High Energy Physics (HEP), including detector responses physics events. However, training GANs notoriously hard optimizing their hyperparameters even more so. normally requires trial-and-error attempts to force stable reach reasonable fidelity. Significant tuning work be done achieve the accuracy...
We present a new software, HYPPO, that enables the automatic tuning of hyperparameters various deep learning (DL) models. Unlike other hyperparameter optimization (HPO) methods, HYPPO uses adaptive surrogate models and directly accounts for uncertainty in model predictions to find accurate reliable make robust predictions. Using asynchronous nested parallelism, we are able significantly alleviate computational burden training complex architectures quantifying uncertainty. is implemented...
Abstract Physical parameterizations in global atmospheric and ocean models typically include free parameters that are not theoretically or empirically constrained. New methods required to determine the optimal parameter combinations for such an objective, exhaustive, yet computationally feasible manner. Here we propose apply inexpensive radial basis function (RBF) surrogate minimize a “cost,” error, of model physical parameterization. The RBF is iteratively updated as more input‐output pairs...