A5090787101- Advanced Database Systems and Queries
- Data Management and Algorithms
- Semantic Web and Ontologies
- Distributed systems and fault tolerance
- Geographic Information Systems Studies
- Interactive and Immersive Displays
- Service-Oriented Architecture and Web Services
- Advanced Computational Techniques and Applications
- Distributed and Parallel Computing Systems
- Tactile and Sensory Interactions
- Advanced Data Storage Technologies
- Augmented Reality Applications
- Hand Gesture Recognition Systems
- Data Mining Algorithms and Applications
- EEG and Brain-Computer Interfaces
- Data Quality and Management
- Dementia and Cognitive Impairment Research
- Functional Brain Connectivity Studies
- Artificial Intelligence in Healthcare
- Algorithms and Data Compression
- Advanced Image and Video Retrieval Techniques
- Automated Road and Building Extraction
- Virtual Reality Applications and Impacts
- Parallel Computing and Optimization Techniques
- Usability and User Interface Design
Florida International University
2016-2025
Texas State University
2023
Shahjalal University of Science and Technology
2023
Technical Database Services (United States)
2020
Miami Dade College
2019
Colorado State University
2019
University of Miami
2009-2014
Florida Atlantic University
2011
Lomonosov Moscow State University
2002-2003
University of Illinois Chicago
2002
We discuss our approach to developing a novel modality for the computer-delivery of Brief Motivational Interventions (BMIs) behavior change in form personalized On-Demand VIrtual Counselor (ODVIC), accessed over internet. ODVIC is multimodal Embodied Conversational Agent (ECA) that empathically delivers an evidence-based intervention by adapting, real-time, its verbal and nonverbal communication messages those user’s during their interaction. currently focus work on excessive alcohol...
Predicting the progression of Alzheimer's Disease (AD) has been held back for decades due to lack sufficient longitudinal data required development novel machine learning algorithms. This study proposes a algorithm predicting disease using distributed multimodal, multitask method. More specifically, each individual task is defined as regression model, which predicts cognitive scores at single time point. Since prediction tasks multiple intervals are related other in chronological order,...
Motivated by the challenge of navigating complex landscape cybersecurity compliance, this study critically examines and evaluates seven major frameworks: SOC 2, GDPR, PCI DSS, HIPAA, CIS Controls V8, NIST CSF, CMMC 2.0. Our research focuses on understanding their distinct features operational nuances, addressing a significant gap in current compliance strategies. We contribute novel set risk management-based evaluation criteria, offering comprehensive analysis these frameworks. The further...
Road extraction is a sub-domain of remote sensing applications; it subject extensive and ongoing research. The procedure automatically extracting roads from satellite imagery encounters significant challenges due to the multi-scale diverse structures roads; improvement in this field needed. Convolutional neural networks (CNNs), especially DeepLab series known for its proficiency semantic segmentation efficiency interpreting objects’ features, address some these caused by varying nature...
To study the neural networks reorganization in pediatric epilepsy, a consortium of imaging centers was established to collect functional data. Common paradigms and similar acquisition parameters were used. We studied 122 children (64 control 58 LRE patients) across five sites using EPI BOLD fMRI an auditory description decision task. After normalization MNI atlas, activation maps generated by FSL separated into three sub-groups distance method principal component analysis (PCA)-based...
Smart cities combine technology and human resources to improve the quality of life reduce expenditures. Ensuring safety city residents remains one open problems, as standard budgetary investments fail decrease crime levels. This work takes steps toward implementing smart, safe cities, by combining use personal mobile devices social networks make users aware their surroundings. We propose novel metrics define location user based values. evaluate ability forecasting techniques including...
In this paper, we discuss a novel approach for the computer-delivery of Brief Motivational Interventions (BMIs) health behavior change. We describe basic elements our system architecture, and focus on enabling multimodal Embodied Conversational Agent (ECA) to deliver change interventions empathetically by adapting, in real-time, its verbal non-verbal communication messages those clients. The designed empathy model integrates cognitive component an affective components. then evaluation...
This paper proposes a novel strategy for estimating the size of resulting relation after an equi-join and selection using regression model. An approximating series representing underlying data distribution dependency is derived from actual data. The proposed method provides instant accurate estimation by performing evaluation series, with no run-time overheads in page faults space, negligible CPU overhead. In contrast, popular sampling methods incur (for sampling), time space. These increase...
With the introduction of new input devices, a series questions have been raised in regard to making user interaction more intuitive - particular, preferred gestures for different tasks. Our study looks into how find gesture set 3D travel using multi-touch display and mid-air device improve interaction. We conducted with 30 subjects, concluding that users simple multi-touch. In addition, we found legacy carried over mid-Air Finally, propose both type interactions.
K-means clustering is one of the most commonly used methods for classification and data-mining. When amount data to be clustered "huge," and/or when becomes available in increments, has devise incremental procedures. Current research on does not address several specific problems including seeding problem, sensitivity algorithm order data, number clusters. In this paper we present static dynamic single-pass procedures that overcome these limitations.
This paper proposes an implementation of Recurrent Neural Networks (RNNs) for (a) predicting future Mini-Mental State Examination (MMSE) scores in a longitudinal study and (b) deploying multiclass multimodal neuroimaging classification process that involves three different known stages Alzheimer's progression, cognitively normal (CN), Mild Cognitive Impairment (MCI) Disease (AD). data is fed into two well-studied variations the RNNs; Long Short-Term Memory (LSTM) Gated Unit (GRU). The...
Generative models have recently gained popularity in Remote Sensing, offering substantial benefits for interpreting and utilizing satellite imagery across diverse applications such as climate monitoring, urban planning, wildfire detection. These are particularly adept at addressing the challenges posed by images, which often exhibit domain variability due to seasonal changes, sensor characteristics, especially variations spectral bands. Such can significantly impact model performance various...
In this work, we leverage Cloud computing technologies in scaling out data management geographical databases. particular, tackle the issue of indexing parallel. First, spatial is partitioned and indexed a Hadoop MapReduce cluster. Two main partitioning strategies are evaluated: a) A linear-complexity method based on Zorder values, b) An iterative algorithm X-means clustering. The advantages drawbacks each weighted with relation to query performance. Second, interactive queries processed from...
The world is facing an epidemic rise in diabetes mellitus (DM) incidence, which challenging health funders, systems, clinicians, and patients to understand respond a flood of research knowledge.Evidence-based guidelines provide uniform management recommendations for "average" that rarely take into account individual variation susceptibility DM, its complications, responses pharmacological lifestyle interventions.Personalized medicine combines bioinformatics with genomic, proteomic,...
A complex fuzzy class is characterized by a pure grade of membership. Pure classes are paramount in providing rich semantics for cases where the data periodic with period. Often, however, available contaminated noise, opposing expert opinions, ambiguity, and false information. This opens door using intuitionistic sets theory: representing information via degree non-membership. Several researchers have identified benefits integrating two concepts sets. Nevertheless, allow only one component...