A5039017664- Protein Structure and Dynamics
- COVID-19 diagnosis using AI
- Topic Modeling
- Radiomics and Machine Learning in Medical Imaging
- Gaussian Processes and Bayesian Inference
- Machine Learning in Healthcare
- Enzyme Structure and Function
- Statistical Methods and Inference
- Machine Learning in Bioinformatics
- Fault Detection and Control Systems
- Biomedical Text Mining and Ontologies
- Machine Learning in Materials Science
- SARS-CoV-2 and COVID-19 Research
- Lung Cancer Diagnosis and Treatment
- Machine Learning and Data Classification
- Advanced Bandit Algorithms Research
- Bacteriophages and microbial interactions
- Respiratory Support and Mechanisms
- Sepsis Diagnosis and Treatment
- Healthcare Policy and Management
- Data Stream Mining Techniques
- Scientific Computing and Data Management
- Neural Networks and Applications
- Geological Modeling and Analysis
- Reinforcement Learning in Robotics
Calumet College of Saint Joseph
2023
Oak Ridge National Laboratory
2017-2022
University of Limerick
1998
New York State Department of Health
1995
Bidirectional Encoder Representations from Transformers (BERT) and BERT-based approaches are the current state-of-the-art in many natural language processing (NLP) tasks; however, their application to document classification on long clinical texts is limited. In this work, we introduce four methods scale BERT, which by default can only handle input sequences up approximately 400 words long, perform several thousand long. We compare these against two much simpler architectures - a word-level...
We explored how a deep learning (DL) approach based on hierarchical attention networks (HANs) can improve model performance for multiple information extraction tasks from unstructured cancer pathology reports compared to conventional methods that do not sufficiently capture syntactic and semantic contexts free-text documents.Data our analyses were obtained 942 deidentified collected by the National Cancer Institute Surveillance, Epidemiology, End Results program. The HAN was implemented 2...
We examine the problem of clustering biomolecular simulations using deep learning techniques. Since simulation datasets are inherently high dimensional, it is often necessary to build low dimensional representations that can be used extract quantitative insights into atomistic mechanisms underlie complex biological processes. use a convolutional variational autoencoder (CVAE) learn biophysically relevant latent features from long time-scale protein folding in an unsupervised manner....
Markov chain Monte Carlo (MCMC) methods have not been broadly adopted in Bayesian neural networks (BNNs). This paper initially reviews the main challenges sampling from parameter posterior of a network via MCMC. Such culminate to lack convergence posterior. Nevertheless, this shows that nonconverged chain, generated MCMC space network, can yield marginalization valuable predictive distribution output network. Classification examples based on multilayer perceptrons showcase highly accurate...
Significant strides have been made in supervised learning settings thanks to the successful application of deep learning. Now, recent work has brought techniques bear on sequential decision processes area reinforcement (DRL). Currently, little is known regarding hyperparameter optimization for DRL algorithms. Given that algorithms are computationally intensive train, and be sample inefficient, optimizing model hyperparameters presents significant challenges established techniques. We provide...
We introduce novel communication strategies in synchronous distributed Deep Learning consisting of decentralized gradient reduction orchestration and computational graph-aware grouping tensors. These new techniques produce an optimal overlap between computation result near-linear scaling (0.93) training up to 27,600 NVIDIA V100 GPUs on the Summit Supercomputer. demonstrate our context a Fully Convolutional Neural Network approximate solution longstanding scientific inverse problem materials...
As machine learning models continue to increase in complexity, so does the potential number of free model parameters commonly known as hyperparameters. While there has been considerable progress toward finding optimal configurations these hyperparameters, many optimization procedures are treated black boxes. We believe methods should not only return a set optimized but also give insight into effects hyperparameter settings. To this end, we present HyperSpace, parallel implementation Bayesian...
Graph-of-words is a flexible and efficient text representation which addresses well-known challenges, such as word ordering variation of expressions, to natural language processing. In this paper, we consider the latest graph-based convolutional neural network technique, Text GraphConvolutional Network (Text GCN), in context performingclassification tasks on free-form texts. To do this, designed study multi-task information extraction from medical documents. We implemented learning GCN,...
Abstract The emergence and rapid worldwide spread of the novel coronavirus disease, COVID-19, has prompted concerted efforts to find successful treatments. causative virus, severe acute respiratory syndrome 2 (SARS-CoV-2), uses its spike (S) protein gain entry into host cells. Therefore, S presents a viable target develop directed therapy. Here, we deployed an integrated artificial intelligence with molecular dynamics simulation approach provide new details structure. Based on comprehensive...
Significant progress in computer hardware and software have enabled molecular dynamics (MD) simulations to model complex biological phenomena such as protein folding. However, enabling MD access biologically relevant timescales (e.g., beyond milliseconds) still remains challenging. These limitations include (1) quantifying which set of states already been (sufficiently) sampled an ensemble runs, (2) identifying novel from can be initiated sample rare events sampling folding events). With the...
The emergence and rapid worldwide spread of the novel coronavirus disease, COVID-19, has prompted concerted efforts to find successful treatments. causative virus, severe acute respiratory syndrome 2 (SARS-CoV-2), uses its spike (S) protein gain entry into host cells. Therefore, S presents a viable target develop directed therapy. Here, we deployed an integrated artificial intelligence with all-atom molecular dynamics simulation approach provide new details structure. Based on comprehensive...
Abstract We examine the problem of clustering biomolecular simulations using deep learning techniques. Since simulation datasets are inherently high dimensional, it is often necessary to build low dimensional representations that can be used extract quantitative insights into atomistic mechanisms underlie complex biological processes. In this paper, we use a convolutional variational autoencoder (CVAE) learn biophysically relevant latent features from long time-scale protein folding in an...
Introduction: Mechanical ventilation is a life-saving treatment in the Intensive Care Unit (ICU), but often causes patients to be at risk of further respiratory complication. We created statistical model utilizing electronic health record and physiologic vitals data predict Center for Disease Control Prevention (CDC) defined Ventilator Associated Complications (VACs). Further, we evaluated effect temporal resolution feature generation method choice on accuracy such constructed model....
Deep learning (DL) models are being deployed at medical centers to aid radiologists for diagnosis of lung conditions from chest radiographs. Such often trained on a large volume publicly available labeled These pre-trained DL models' ability generalize in clinical settings is poor because the changes data distributions between and privately held In radiographs, heterogeneity arises diverse X-ray equipment their configurations used generating images. machine community, challenges posed by...
Deep learning (DL) models are being deployed at medical centers to aid radiologists for diagnosis of lung conditions from chest radiographs. Such often trained on a large volume publicly available labeled These pre-trained DL models' ability generalize in clinical settings is poor because the changes data distributions between and privately held In radiographs, heterogeneity arises diverse X-ray equipment their configurations used generating images. machine community, challenges posed by...
The use of AI and ML for scientific applications is currently a very exciting dynamic field. Much this excitement HPC has focused on whose analysis classification generate large numbers flops. Others seek to replace simulations with data-driven surrogate models. But another important case lies in the combination application improve simulation accuracy. To that end, we present an anomaly problem which arises from core-collapse supernovae simulation. We discuss strategies early successes...
Deep learning (DL) models are being deployed at medical centers to aid radiologists for diagnosis of lung conditions from chest radiographs. Such often trained on a large volume publicly available labeled These pre-trained DL models' ability generalize in clinical settings is poor because the changes data distributions between and privately held In radiographs, heterogeneity arises diverse X-ray equipment their configurations used generating images. machine community, challenges posed by...