A5045267281- Anomaly Detection Techniques and Applications
- Context-Aware Activity Recognition Systems
- Dementia and Cognitive Impairment Research
- Human Pose and Action Recognition
- Technology Use by Older Adults
- Online Learning and Analytics
- Advanced Clustering Algorithms Research
- Non-Invasive Vital Sign Monitoring
- Gait Recognition and Analysis
- Data Management and Algorithms
- Geriatric Care and Nursing Homes
- Diabetic Foot Ulcer Assessment and Management
- Face and Expression Recognition
- Aging and Gerontology Research
- Stroke Rehabilitation and Recovery
- Health disparities and outcomes
- EEG and Brain-Computer Interfaces
- Emotion and Mood Recognition
- Imbalanced Data Classification Techniques
- COVID-19 diagnosis using AI
- Intelligent Tutoring Systems and Adaptive Learning
- Machine Learning in Healthcare
- Artificial Intelligence in Healthcare and Education
- Mind wandering and attention
- Text and Document Classification Technologies
University Health Network
2017-2025
University of Toronto
2017-2025
Toronto Rehabilitation Institute
2017-2025
American University of the Middle East
2024-2025
New York University
2025
Jinnah Medical & Dental College
2023-2024
Karachi Medical and Dental College
2023-2024
Health Net
2019-2024
University of Georgia
2023
University of Manitoba
2023
One-class classification (OCC) algorithms aim to build models when the negative class is either absent, poorly sampled or not well defined. This unique situation constrains learning of efficient classifiers by defining boundary just with knowledge positive class. The OCC problem has been considered and applied under many research themes, such as outlier/novelty detection concept learning. In this paper we present a unified view general presenting taxonomy study for problems, which based on...
Mixed data comprises both numeric and categorical features, mixed datasets occur frequently in many domains, such as health, finance, marketing. Clustering is often applied to find structures group similar objects for further analysis. However, clustering challenging because it difficult directly apply mathematical operations, summation or averaging, the feature values of these datasets. In this paper, we present a taxonomy study algorithms by identifying five major research themes. We then...
Abstract Artificial intelligence (AI) and machine learning are changing our world through their impact on sectors including health care, education, employment, finance, law. AI systems developed using data that reflect the implicit explicit biases of society, there significant concerns about how predictive models in amplify inequity, privilege, power society. The widespread applications have led to mainstream discourse perpetuating racism, sexism, classism; yet, ageism been largely absent...
Behavioral and psychological symptoms of dementia (BPSD) signal distress or unmet needs present a risk to people with their caregivers. Variability in the expression these is barrier performance digital biomarkers. The aim this study was use wearable multimodal sensors develop personalized machine learning models capable detecting individual patterns BPSD.Older adults BPSD (n = 17) on care unit wore wristband during waking hours for up 8 weeks. captured motion (accelerometer) physiological...
Abstract Background The escalating impact of diabetes and its complications, including diabetic foot ulcers (DFUs), presents global challenges in quality life, economics, resources, affecting around half a billion people. DFU healing is hindered by hyperglycemia-related issues diverse diabetes-related physiological changes, necessitating ongoing personalized care. Artificial intelligence clinical research strive to address these facilitating early detection efficient treatments despite...
The existing methods for video anomaly detection mostly utilize videos containing identifiable facial and appearance-based features. use of with faces raises privacy concerns, especially when used in a hospital or community-based setting. Appearance-based features can also be sensitive to pixel-based noise, straining the model changes background making it difficult focus on actions humans foreground. Structural information form skeletons describing human motion is privacy-protecting overcome...
Automatic detection of students' engagement in online learning settings is a key element to improve the quality and deliver personalized materials them. Varying levels exhibited by students an classroom affective behavior that takes place over space time. Therefore, we formulate detecting from videos as spatio-temporal classification problem. In this paper, present novel end-to-end Residual Network (ResNet) Temporal Convolutional (TCN) hybrid neural network architecture for level videos. The...
The COVID-19 pandemic has exposed persistent inequities in the long-term care sector and brought strict social/physical distancing public health quarantine guidelines that inadvertently put residents at risk for social isolation loneliness. Virtual communication technologies have come to forefront as primary mode maintain connections with their loved ones outside world; yet, many homes do not technological capabilities support modern day technologies. There is an urgent need replace...
Behavioural symptoms of dementia present a significant risk within Long Term Care (LTC) homes, which face difficulties supporting residents and monitoring their safety with limited staffing resources. Many LTC facilities have installed video surveillance systems in common areas that can help staff to observe residents; however, typically these streams are not monitored. In this paper, we the development computer vision algorithm use detect episodes clinically important agitation people...
In one-class classification problems, only the data for target class is available, whereas non-target may be completely absent. this paper, we study nearest neighbor (OCNN) classifiers and their different variants. We present a theoretical analysis to show relationships among variants of OCNN that use neighbors or thresholds identify unseen examples class. also method based on inter-quartile range optimizing parameters used in absence during training. Then, propose two ensemble approaches...