A5048176207- Quantum Computing Algorithms and Architecture
- Anomaly Detection Techniques and Applications
- Scientific Computing and Data Management
- Solar Radiation and Photovoltaics
- Quantum Information and Cryptography
- Machine Learning in Materials Science
- Computational Drug Discovery Methods
- Stochastic Gradient Optimization Techniques
- Distributed and Parallel Computing Systems
- Bioinformatics and Genomic Networks
- Advanced X-ray Imaging Techniques
- Advanced Electron Microscopy Techniques and Applications
- Particle physics theoretical and experimental studies
- Time Series Analysis and Forecasting
- Cloud Computing and Resource Management
- Functional Brain Connectivity Studies
- Photovoltaic System Optimization Techniques
- Particle Detector Development and Performance
- Meteorological Phenomena and Simulations
- Advanced Neural Network Applications
- Electron and X-Ray Spectroscopy Techniques
- Generative Adversarial Networks and Image Synthesis
- Advancements in Semiconductor Devices and Circuit Design
- Advanced Data Storage Technologies
- Adversarial Robustness in Machine Learning
Brookhaven National Laboratory
2016-2025
University of California, Davis
2025
Catholic University of Korea
2024
ExxonMobil (United States)
2023-2024
Sandia National Laboratories
2023-2024
Robert Bosch (Germany)
2023-2024
University of Maryland, College Park
2023-2024
RIKEN Center for Advanced Photonics
2023-2024
RIKEN Center for Quantum Computing
2023-2024
Computational Physics (United States)
2021-2022
To the Editor: Over past two decades, scale and complexity of genomics technologies data have advanced from sequencing genomes a few organisms to generating metagenomes, genome variation, gene expression, metabolites, phenotype for thousands their communities.A major challenge in this data-rich age biology is integrating heterogeneous distributed into predictive models biological function, ranging single entire ecologies.The US Department Energy (DOE) has invested substantially efforts...
We present a supercomputer-driven pipeline for in silico drug discovery using enhanced sampling molecular dynamics (MD) and ensemble docking. Ensemble docking makes use of MD results by compound databases into representative protein binding-site conformations, thus taking account the dynamic properties binding sites. also describe preliminary obtained 24 systems involving eight proteins proteome SARS-CoV-2. The involves temperature replica exchange sampling, making massively parallel...
Long short-term memory (LSTM) is a kind of recurrent neural networks (RNN) for sequence and temporal dependency data modeling its effectiveness has been extensively established. In this work, we propose hybrid quantum-classical model LSTM, which dub QLSTM. We demonstrate that the proposed successfully learns several kinds data. particular, show certain testing cases, quantum version LSTM converges faster, or equivalently, reaches better accuracy, than classical counterpart. Due to...
Unpaired image-to-image translation has broad applications in art, design, and scientific simulations. One early breakthrough was CycleGAN that emphasizes one-to-one mappings between two unpaired image domains via generative-adversarial networks (GAN) coupled with the cycle-consistency constraint, while more recent works promote one-to-many mapping to boost diversity of translated images. Motivated by simulation needs, this work revisits classic framework boosts its performance outperform...
This paper presents a quantum convolutional neural network (QCNN) for the classification of high energy physics events. The proposed model is tested using simulated dataset from Deep Underground Neutrino Experiment. architecture demonstrates an advantage learning faster than classical networks (CNNs) under similar number parameters. In addition to convergence, QCNN achieves greater test accuracy compared CNNs. Based on our results numerical simulations, it promising direction apply and other...
Quantum computers offer an intriguing path for a paradigmatic change of computing in the natural sciences and beyond, with potential achieving so-called quantum advantage—namely, significant (in some cases exponential) speedup numerical simulations. The rapid development hardware devices various realizations qubits enables execution small-scale but representative applications on computers. In particular, high-energy physics community plays pivotal role accessing power computing, since field...
Abstract Third-generation long-range DNA sequencing and mapping technologies are creating a renaissance in high-quality genome sequencing. Unlike second-generation sequencing, which produces short reads few hundred base-pairs long, third-generation single-molecule generate over 10,000 bp or map 100,000 molecules. We analyze how increased read lengths can be used to address longstanding problems de novo assembly, structural variation analysis haplotype phasing.
We use quantum detector tomography to characterize the qubit readout in terms of measurement positive operator-valued measures (POVMs) on IBM computers Q 5 Tenerife and Yorktown. Our results suggest that characterized model deviates from ideal projectors, ranging 10 40%. This is mostly dominated by classical errors, evident shrinkage arrows poles corresponding Bloch-vector representations. There are also small deviations not ``classical,'' order 3% or less, represented tilt $z$ axis. Further...
Abstract Third generation single molecule sequencing technology is poised to revolutionize genomics by enabling the of long, individual molecules DNA and RNA. These technologies now routinely produce reads exceeding 5,000 basepairs, can achieve as long 50,000 basepairs. Here we evaluate limits assessing impact read in assembly human genome 25 other important genomes across tree life. From this, develop a new data-driven model using support vector regression that accurately predict...
Abstract Nationwide population-based cohort provides a new opportunity to build an automated risk prediction model based on individuals’ history of health and healthcare beyond existing models. We tested the possibility machine learning models predict future incidence Alzheimer’s disease (AD) using large-scale administrative data. From Korean National Health Insurance Service database between 2002 2010, we obtained de-identified data in elders above 65 years ( N = 40,736) containing 4,894...
Distributed training across several quantum computers could significantly improve the time and if we share learned model, not data, it potentially data privacy as would happen where is located. One of potential schemes to achieve this property federated learning (FL), which consists clients or local nodes on their own a central node aggregate models collected from those nodes. However, best our knowledge, no work has been done in machine (QML) federation setting yet. In work, present hybrid...
Simulations of excited state properties, such as spectral functions, are often computationally expensive and therefore not suitable for high-throughput modeling. As a proof principle, we demonstrate that graph-based neural networks can be used to predict the x-ray absorption near-edge structure spectra molecules quantitative accuracy. Specifically, predicted reproduce nearly all prominent peaks, with 90% peak locations within 1 eV ground truth. Besides its own utility in analysis inference,...
Quantum machine learning could possibly become a valuable alternative to classical for applications in High Energy Physics by offering computational speed-ups. In this study, we employ support vector with quantum kernel estimator (QSVM-Kernel method) recent LHC flagship physics analysis: $t\bar{t}H$ (Higgs boson production association top quark pair). our simulation study using up 20 qubits and 50000 events, the QSVM-Kernel method performs as well its counterparts three different platforms...
One of the major objectives experimental programs at Large Hadron Collider (LHC) is discovery new physics. This requires identification rare signals in immense backgrounds. Using machine learning algorithms greatly enhances our ability to achieve this objective. With progress quantum technologies, could become a powerful tool for data analysis high energy In study, using IBM gate-model computing systems, we employ variational classifier method two recent LHC flagship physics analyses: (Higgs...
Abstract Quantum machine learning (QML) can complement the growing trend of using learned models for a myriad classification tasks, from image recognition to natural speech processing. There exists potential quantum advantage due intractability operations on classical computer. Many datasets used in are crowd sourced or contain some private information, but best our knowledge, no current QML equipped with privacy-preserving features. This raises concerns as it is paramount that do not expose...
Quantum computers offer an intriguing path for a paradigmatic change of computing in the natural sciences and beyond, with potential achieving so-called quantum advantage, namely significant (in some cases exponential) speed-up numerical simulations. The rapid development hardware devices various realizations qubits enables execution small scale but representative applications on computers. In particular, high-energy physics community plays pivotal role accessing power computing, since field...
Email is one of the most prevalent communication tools today, and solving email overload problem pressingly urgent. A good way to alleviate automatically prioritize received messages according priorities each user. However, research on statistical learning methods for fully personalized prioritization (PEP) has been sparse due privacy issues, since people are reluctant share personal importance judgments with community. It therefore important develop evaluate PEP under assumption that only...
Carlos Soto, Shinjae Yoo. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint (EMNLP-IJCNLP). 2019.
Abstract As a critical component of coherent X-ray diffraction imaging (CDI), phase retrieval has been extensively applied in structural science to recover the 3D morphological information inside measured particles. Despite meeting all oversampling requirements Sayre and Shannon, current approaches still have trouble achieving unique inversion experimental data presence noise. Here, we propose overcome this limitation by incorporating Machine Learning (ML) model combining (optional)...
Domain Generalization (DG) aims to learn a generalizable model on the unseen target domain by only training multiple observed source domains. Although variety of DG methods have focused extracting domain-invariant features, domain-specific class-relevant features attracted attention and been argued benefit generalization domain. To take into account information, in this paper we propose an Information theory iNspired diSentanglement pURification modEl (INSURE) explicitly disentangle latent...
Abstract Quantum federated learning (QFL) can facilitate collaborative across multiple clients using quantum machine (QML) models, while preserving data privacy. Although recent advances in QFL span different tasks like classification leveraging several types, no prior work has focused on developing a framework that utilizes temporal to approximate functions useful analyze the performance of distributed sensing networks. In this paper, novel is first integrate long short-term memory (QLSTM)...
Abstract We present an ensemble transfer learning method to predict suicide from Veterans Affairs (VA) electronic medical records (EMR). A diverse set of base models was trained a binary outcome constructed reported suicide, attempt, and overdose diagnoses with varying choices study design prediction methodology. Each model used twenty cross-sectional 190 longitudinal variables observed in eight time intervals covering 7.5 years prior the prediction. Ensembles seven were created fine-tuned...