A5081968453- Bioinformatics and Genomic Networks
- Multimodal Machine Learning Applications
- Dementia and Cognitive Impairment Research
- Software Engineering Research
- Stock Market Forecasting Methods
- Satellite Image Processing and Photogrammetry
- Advanced Image and Video Retrieval Techniques
- Software Reliability and Analysis Research
- Chronic Disease Management Strategies
- Advanced Malware Detection Techniques
- Reservoir Engineering and Simulation Methods
- Anomaly Detection Techniques and Applications
- Mental Health Research Topics
- Caching and Content Delivery
- Subtitles and Audiovisual Media
- Optimization and Search Problems
- Human Pose and Action Recognition
- Lung Cancer Diagnosis and Treatment
- Complex Systems and Time Series Analysis
- Advanced Graph Neural Networks
- Radiomics and Machine Learning in Medical Imaging
- Energy Load and Power Forecasting
- Opportunistic and Delay-Tolerant Networks
- Explainable Artificial Intelligence (XAI)
- Advanced Radiotherapy Techniques
The University of Texas at San Antonio
2019-2024
Recurrent neural networks have received vast amount of attention in time series prediction due to their flexibility capturing dependencies on various scales. However, as most the classical forecasting methods, its accuracy is strongly tied degree signal complexity. Specifically, stock market prices are commonly classified be non-linear, non-stationary and chaotic signals, since they exhibit erratic behavior that conducts a poor performance long short-term memory (LSTM). In this paper, we...
A great number of computer vision publications have focused on distinguishing between human action recognition and classification rather than the intensity actions performed. Indexing which determines performance is a challenging task due to uncertainty information deficiency that exists in video inputs. To remedy this uncertainty, paper we coupled fuzzy logic rules with neural-based model rate as intense or mild. In our approach, used Spatio-Temporal LSTM generate weights fuzzy-logic model,...
Graphs in real-world applications are dynamic both terms of structures and inputs. Information discovery such networks, which present dense deeply connected patterns locally sparsity globally can be time consuming computationally costly. In this paper we address the shortest path query spatio-temporal graphs is a fundamental graph problem with numerous applications. graphs, classical algorithms insufficient or even flawed because information consistency not guaranteed between two timestamps...
It is important but challenging to understand the interactions of multiple chronic conditions (MCC) and how they develop over time in patients populations. Clinical data on MCC can now be represented using graphical models study their interaction identify path toward development MCC. However, current representing are often complex difficult analyze. Therefore, it necessary improved methods for generating these models.This aimed summarize improve comprehension aid analysis.We examined...
With growing applications of Machine Learning (ML) techniques in the real world, it is highly important to ensure that these models work an equitable manner. One main step ensuring fairness effectively measure fairness, and this end, various metrics have been proposed past literature. While computation those are straightforward classification set-up, computationally intractable regression domain. To address challenge computational intractability, literature methods approximate such metrics....
Over the years, open-source software systems have become prey to threat actors. Even as communities act quickly patch breach, code vulnerability screening should be an integral part of agile development from beginning. Unfortunately, current techniques are ineffective at identifying novel vulnerabilities or providing developers with and classification. Furthermore, datasets used for learning often exhibit distribution shifts real-world testing due attack strategies deployed by adversaries a...
<sec> <title>BACKGROUND</title> It is important but challenging to understand the interactions of multiple chronic conditions (MCC) and how they develop over time in patients populations. Clinical data on MCC can now be represented using graphical models study their interaction identify path toward development MCC. However, current representing are often complex difficult analyze. Therefore, it necessary improved methods for generating these models. </sec> <title>OBJECTIVE</title> This aimed...