A5089449834- Frailty in Older Adults
- Semantic Web and Ontologies
- Biomedical Text Mining and Ontologies
- Public Relations and Crisis Communication
- Tropical and Extratropical Cyclones Research
- Advanced Text Analysis Techniques
- Disaster Management and Resilience
- Cryptography and Data Security
- Cloud Data Security Solutions
- Anomaly Detection Techniques and Applications
- Data-Driven Disease Surveillance
- Health Systems, Economic Evaluations, Quality of Life
- Chronic Disease Management Strategies
- Topic Modeling
- Internet Traffic Analysis and Secure E-voting
- Security and Verification in Computing
- Hip and Femur Fractures
- Flood Risk Assessment and Management
- Sentiment Analysis and Opinion Mining
- Cardiac, Anesthesia and Surgical Outcomes
- Artificial Intelligence in Healthcare
- Earthquake Detection and Analysis
- Natural Language Processing Techniques
- Heart Failure Treatment and Management
- Geriatric Care and Nursing Homes
UnitedHealth Group (United States)
2023-2025
University of North Carolina at Greensboro
2016-2024
Maulana Azad National Institute of Technology
2024
Optum (United States)
2023
Mississippi State University
2012-2015
Diabetes and cardiovascular disease are two of the main causes death in United States. Identifying predicting these diseases patients is first step towards stopping their progression. We evaluate capabilities machine learning models detecting at-risk using survey data (and laboratory results), identify key variables within contributing to among patients.Our research explores data-driven approaches which utilize supervised with such diseases. Using National Health Nutrition Examination Survey...
Healthcare costs due to unplanned readmissions are high and negatively affect health wellness of patients. Hospital readmission is an undesirable outcome for elderly Here, we present risk prediction using five machine learning approaches predicting 30-day patients (age ≥ 50 years). We use a comprehensive curated set variables that include frailty, comorbidities, high-risk medications, demographics, hospital, insurance utilization build these models. conduct large-scale study with electronic...
Ontologies are critical for organizing and interpreting complex domain-specific knowledge, with applications in data integration, functional prediction, knowledge discovery. As the manual curation of ontology annotations becomes increasingly infeasible due to exponential growth biomedical genomic data, natural language processing (NLP)-based systems have emerged as scalable alternatives. Evaluating these requires robust semantic similarity metrics that account hierarchical partially correct...
There is a growing interest in using social media content for Natural Language Processing applications. However, it not easy to computationally identify the most relevant set of tweets related any specific event. Challenging semantics coupled with different ways natural language make difficult retrieving data from outlet. This paper seeks demonstrate way present changing Twitter within context crisis event, specifically during Hurricane Irma. These methods can be used corpus text analysis...
Predicting human mobility within cities is an important task in urban and transportation planning. With the vast amount of digital traces available through social media platforms, we investigate potential application such data predicting commuter trip distribution at small spatial scale. We develop back propagation (BP) neural network gravity models using both traditional Twitter New York City to explore their performance compare results. Our results suggest modeling improve prediction...
Abstract Classifying images using supervised machine learning (ML) relies on labeled training data—classes or text descriptions, for example, associated with each image. Data‐driven models are only as good the data used training, and this points to importance of high‐quality developing a ML model that has predictive skill. Labeling is typically time‐consuming, manual process. Here, we investigate process labeling data, specific focus coastal aerial imagery captured in wake hurricanes...
This work presents an on-device machine learning model with the ability to identify different mobility gestures called human activity recognition (HAR), which includes running, walking, squatting, jumping, and others. The data is collected through Arduino Nano 33 BLE Sense board a sampling rate of 119 Hz, embedded Inertial Measurement Unit (IMU) sensor. same used as microcontroller by developing end-to-end edge computing application. A deep neural network trained then compressed for...
This imagery aids in recovery efforts as well rapid assessment of storm impacts along developed and undeveloped coastlines (Madore, Imahori, Kum, White, & Worthem, 2018).
Abstract Emerging adults (EAs) are at high risk for mental health challenges and frequently reach out to their parents support. Yet little is known about how help emerging manage cope with daily stressors which strategies hinder EA health. In this cross‐sectional pilot study of students a 2‐ 4‐year college (ages 18–25, N = 680, mean age 19.0), we extend models dyadic coping from intimate relationships the parent‐emerging adult relationship test whether six specific parent stress associated...
Representing scientific knowledge using ontologies enables data integration, consistent machine-readable representation, and allows for large-scale computational analyses. Text mining approaches that can automatically process annotate literature with ontology concepts are necessary to keep up the rapid pace of publishing. Here, we present deep learning models (Gated Recurrent Units (GRU) Long Short Term Memory (LSTM)) combined different input encoding formats automated Named Entity...
The purpose of the current study was to investigate predictive properties five definitions a frailty risk score (FRS) and three comorbidity indices using data from electronic health records (EHRs) hospitalized adults aged ≥50 years for 3-day, 7-day, 30-day readmission, identify an optimal model FRS combination. Retrospective analysis EHR dataset performed, multivariable logistic regression area under curve (AUC) were used examine readmission comorbidity. sample ( N = 55,778) mostly female...
I. Abstract Text mining approaches for automated ontology-based curation of biological and biomedical literature have largely focused on syntactic lexical analysis along with machine learning. Recent advances in deep learning shown increased accuracy textual data annotation. However, the application is a relatively new area prior work has limited set models. Here, we introduce model/architecture based combining multiple Gated Recurrent Units (GRU) character+word input. We use from five...
We introduce a family of authenticated data structures -Ordered Merkle Trees (OMT) -and illustrate their utility in security kernels for wide variety sub-systems.Specifically, the two types OMTs: a) index ordered merkle tree (IOMT) and b) range (ROMT), are investigated suitability various sub-systems Border Gateway Protocol (BGP), Internet's inter-autonomous system routing infrastructure.We outline simple generic kernel functions to maintain OMTs, sub-system specific functionality BGP...
We present an active learning pipeline to identify hurricane impacts on coastal landscapes. Previously unlabeled post-storm images are used in a three component workflow — first online interface is crowd-source labels for imagery; second, convolutional neural network trained using the labeled images; third, model predictions displayed interactive map. Both labeler and map allow scientists provide additional that will be develop large dataset, refined model, improved impact assessments.
Devices participating in mobile ad hoc networks (MANET) are expected to strictly adhere a uniform routing protocol route data packets among themselves. Unfortunately, MANET devices, composed of untrustworthy software and hardware components, expose large attack surface. This can be exploited by attackers gain control over one or more wreak havoc on the subnet. The approach presented this paper secure MANETs restricts surface single module devices trusted (TMM). TMMs deliberately constrained...
With a growing increase in botnet attacks, computer networks are constantly under threat from attacks that cripple cyber-infrastructure. Detecting these real-time proves to be difficult and resource intensive task. One of the pertinent methods detect such is signature based detection using machine learning models. This paper explores efficacy models at detecting data captured large-scale network attacks. Our study provides comprehensive overview performance characteristics two --- Random...
Annotating scientific literature with ontology concepts is a critical task in biology and several other domains for knowledge discovery. Ontology based annotations can power large-scale comparative analyses wide range of applications ranging from evolutionary phenotypes to rare human diseases the study protein functions. Computational methods that tag text terms have included lexical/syntactic methods, traditional machine learning, most recently, deep learning. Here, we present state art...
Healthcare costs that can be attributed to unplanned readmissions are staggeringly high and negatively impact health wellness of patients. In the United States, hospital systems care providers have strong financial motivations reduce in accordance with several government guidelines. One critical steps reducing is recognize factors lead readmission correspondingly identify at-risk patients based on these factors. The availability large volumes electronic records make it possible develop...