A5101739447- Machine Learning in Healthcare
- Data Quality and Management
- Handwritten Text Recognition Techniques
- Artificial Intelligence in Healthcare
- Business Process Modeling and Analysis
- Multimodal Machine Learning Applications
- Human Motion and Animation
- Hand Gesture Recognition Systems
- Natural Language Processing Techniques
- Electronic Health Records Systems
- Advanced Statistical Process Monitoring
- Digital and Cyber Forensics
- Explainable Artificial Intelligence (XAI)
- Healthcare Technology and Patient Monitoring
Rutgers, The State University of New Jersey
2021-2024
Process data constructed from event logs provides valuable insights into procedural dynamics over time. The confidential information in process data, together with the data’s intricate nature, makes datasets not sharable and challenging to collect. Consequently, research is limited using analytics mining domain. In this study, we introduced a synthetic generation task address limitation of data. We generative adversarial network, called ProcessGAN, generate activity sequences corresponding...
This paper describes our system used in SemEval-2023 Task-1: Visual Word Sense Disambiguation (VWSD). The VWSD task is to identify the correct image that corresponds an ambiguous target word given limited textual context. To reduce ambiguity and enhance selection, we proposed several text augmentation techniques, such as prompting, WordNet synonyms, generation. We experimented with different vision-language pre-trained models capture joint features of augmented image. Our approach achieved...
Process data with confidential information cannot be shared directly in public, which hinders the research process mining and analytics. Data encryption methods have been studied to protect data, but they still may decrypted, leads individual identification. We experimented different models of representation learning used learned model generate synthetic data. introduced an adversarial generative network for generation (ProcessGAN) two Transformer networks generator discriminator. evaluated...
Limb stability refers to the ability of limbs maintain posture during physical movements. Observing limb movements and quantizing is crucial for tremor detection. In this paper, we propose a quantitative metric quantify limbs. We applied Otsu threshold segmentation method features handwriting images that were digitized identified. Then employed eight-connected components deviation degree, obtained two parameters can quantitatively evaluate patient's left right hand's stability. Further...