- COVID-19 epidemiological studies
- Emotion and Mood Recognition
- Big Data and Business Intelligence
- Data-Driven Disease Surveillance
- Network Security and Intrusion Detection
- Complex Network Analysis Techniques
- Anomaly Detection Techniques and Applications
- Human Mobility and Location-Based Analysis
- Distributed and Parallel Computing Systems
- Advanced Text Analysis Techniques
- COVID-19 Pandemic Impacts
- IoT and Edge/Fog Computing
- COVID-19 and Mental Health
- Scientific Computing and Data Management
- Cloud Computing and Resource Management
- Solar Radiation and Photovoltaics
- COVID-19 Digital Contact Tracing
- Service-Oriented Architecture and Web Services
- Non-Invasive Vital Sign Monitoring
- Data Stream Mining Techniques
- Geographic Information Systems Studies
- Meteorological Phenomena and Simulations
- Spatial and Panel Data Analysis
- Vehicular Ad Hoc Networks (VANETs)
- Distributed systems and fault tolerance
University of Louisiana at Lafayette
2014-2023
Tampere University
2023
Institute of Informatics of the Slovak Academy of Sciences
2018
University of South Alabama
2017
Abstract Advances in wearable technologies provide the opportunity to monitor many physiological variables continuously. Stress detection has gained increased attention recent years, mainly because early stress can help individuals better manage health minimize negative impacts of long-term exposure. This paper provides a unique dataset created natural working environment hospital. is collection biometric data nurses during COVID-19 outbreak. Studying work complex due social, cultural, and...
The lack of security in today's in-vehicle network make connected vehicles vulnerable to many types cyber-attacks. Replay-based injection attacks are one the easiest type denial-of-service where attacker floods with malicious traffic intent alter vehicle's normal behavior. may exploit this vulnerability launch targeted low-rate which difficult detect because during looks like regular traffic. In paper, we propose a sequence mining approach Control Area Network (CAN). We discuss four...
Power grid operators rely on solar irradiance forecasts to manage uncertainty and variability associated with power. Meteorological factors such as cloud cover, wind direction, speed affect are a high degree of uncertainty. Statistical models fail accurately capture the dependence between these irradiance. In this paper, we introduce idea applying multivariate Gated Recurrent Units (GRU) forecast Direct Normal Irradiance (DNI) hourly. The proposed GRU-based forecasting method is evaluated...
Software Defined Network (SDN) has become one of the most preferred solutions for management large-scale complex networks. The network policies in are difficult to embed on entire devices simultaneously, whereas SDN these can be embedded top network. is divided into two parts which vertically integrated form Currently many aspects classical architecture Internet etched stone – a so-called ossification led major obstacles IPv6 deployment and difficulty using IP multicast services. Yet, there...
The financial sector has been at heightened risk from cyber-attacks in recent years, which further highlights the need for secure methods of authentication. Single-factor authentication tools are not enough to protect sensitive information and systems unauthorized access. In this paper, we present a novel cryptographic model dual using 2FA modern framework strengthen security measures transactions. particular, new methodology brings together symmetric asymmetric encryption schemes with...
Social Media generates information about news and events in real-time. Given the vast amount of data available rate propagation, reliably identifying is a challenge. Most state-of-the-art techniques are post hoc that detect an event after it happened. Our goal to onset as happening using user-generated from Twitter streams. To achieve this goal, we use discriminative model identify change pattern conversations over time. We topic evolution find credible eliminate random noise prevalent many...
Information about events happening in the real world are generated online on social media real-time. There is substantial research done to detect these using information posted websites like Twitter, Tumblr, and Instagram. The depends type of platform website relies upon, such as short messages, pictures, long form articles. In this paper, we extend an existing real-time event detection at onset approach include multiple websites. We present three different approaches merging from two...
Adverse drug events (ADEs) are among the leading causes of death in United States. Although many ADEs detected during pharmaceutical development and FDA approval process, all possible reactions cannot be identified this period. Currently, post-consumer surveillance relies on voluntary reporting systems, such as FDA's Event Reporting System (AERS). With an increase availability medical resources health related data online, interest mining has grown rapidly. This information coupled with...
Abstract Human mobility plays an important role in the dynamics of infectious disease spread. Evidence from initial nationwide lockdowns for COVID− 19 indicates that restricting human is effective strategy to contain While a direct correlation was observed early on, it not known how impacted infection growth rates once are lifted, primarily due modulation by other factors such as face masks, social distancing, and non-linear patterns both growth. This paper introduces piece-wise approach...
Travel patterns and mobility affect the spread of infectious diseases like COVID-19. However, we do not know to what extent local vs. visitor affects growth in number cases. This study evaluates impact state-level understanding with respect cases for COVID United States between March 1, 2020, December 31, 2020. Two metrics, namely transmission risk, were extracted from data capture potential COVID-19 through mobility. A combination three factors: current cases, are used model future using...
Solar irradiance is the measurement of amount power from Sun per unit area. has a very high degree variability, due to many environmental factors, including cloud cover, relative humidity, and air temperature. Predicting solar useful for measuring future energy production scheduling. Real-time can be forecasted using either maching learning or physics-based models, both having their own respective trade-offs. To overcome limitations machine simulations, we propose novel framework in...
Abstract Containing the COVID-19 pandemic while balancing economy has proven to be quite a challenge for world. We still have limited understanding of which combination policies been most effective in flattening curve; given challenges dynamic and evolving nature pandemic, lack quality data etc. This paper introduces novel mining-based approach understand effects different non-pharmaceutical interventions containing infection rate. used association rule mining perform descriptive on publicly...
Affective computing has garnered researchers' attention and interest in recent years as there is a need for AI systems to better understand react human emotions. However, analyzing emotions, such mood or stress, quite complex. While various stress studies use facial expressions wearables, most existing datasets rely on processing data from single modality. This paper presents EmpathicSchool, novel dataset that captures the associated physiological signals, heart rate, electrodermal activity,...
Link prediction refers to estimating the likelihood of a link appearing in future based on current status graph. problem applications various domains such as bioinformatics, social network analysis, cybersecurity and e-commerce. Some these graphs are massive constantly evolving. Many require graph streams be processed them real-time, predict most recent information features may change over time. Existing approaches process large for is non-trivial due following reasons: 1) Graphs required...
Controller Area Network bus systems within ve-hicular networks are not equipped with the tools necessary to ward off and protect themselves from modern cyber-security threats. Work has been done on using machine learning methods detect report these attacks, but common robust towards unknown attacks. These usually rely there being a sufficient representation of attack data, which may be available due either enough data present adequately represent its distribution or itself is too diverse in...
Spatial scan statistics is one of the most important models in order to detect high activity or hotspots real world applications such as epidemiology, public health, astronomy and criminology on geographic data. Traditional statistic uses regular shapes like circles areas activity; same model was extended eclipses improve model. More recent works identify irregular shaped for data with geographical boundaries, where information about population within boundaries available. With introduction...
Processing high-volume, high-velocity data streams is an important big problem in many sciences, engineering, and technology domains. There are open-source distributed stream processing cloud platforms that offer low-latency at scale, but the visualization user-interaction components of these systems limited to visualizing outcome results. Visual analysis represents a new form where user has more control interactive capabilities either dynamically change visualization, analytics or...
Social media plays an important role in communication between people recent times. This includes information about news and events that are currently happening. Most of the research on event detection concentrates identifying from social information. These models assume to be a single entity treat it as such during process. assumption ignores composition changes new is made available media. To capture change over time, we extend already existing Event Detection at Onset algorithm study...