top of page

 

Research Thrust 1: A Laplacian Regularized Graph Neural Network for Predictive Modeling of Multiple Chronic Conditions

"A Laplacian Regularized Graph Neural Network for Predictive Modeling of Multiple Chronic Conditions" introduces a novel approach to analyze the interconnections between various chronic conditions and patient-level risk factors using Graph Neural Networks (GNNs) enhanced with Laplacian regularization. This study aims to overcome the limitations of traditional healthcare analytics by leveraging the structural information of healthcare data through the graph-based model, thereby improving the accuracy and reliability of predicting patient outcomes across multiple chronic conditions. The proposed model demonstrates a significant improvement in predictive performance compared to baseline models, emphasizing the importance of graph structural information in healthcare analytics. This research not only contributes to the advancement of predictive modeling techniques in the domain of chronic diseases but also highlights the potential of graph neural networks in extracting meaningful insights from complex healthcare datasets.

Highlights:

  • Utilizing Laplacian regularized graph neural network to predict multiple chronic conditions using electronic health record data.

  • Enhancing parameter learning by harnessing the multiple chronic conditions network structure.

  • Proven efficacy of Laplacian regularization in graph neural networks for improved predictive   performance.

 

Picture1.png

 

Research Thrust 2: Active Learning Methodology for Design and Optimization of Complex Expensive Tests

In an increasing number of cases involving estimation of a response surface, one is often confronted with situations where there are several factors to be evaluated but experiments are prohibitively expensive. In such scenarios, learning algorithms can actively query the user or other resources to determine the most informative settings to be tested. This area of my research aims at exploring active learning methodologies to minimize the data required for design and optimization of expensive tests for statistical estimation of complex systems.

  • We investigate the fundamental idea of adding a similarity learning penalty to design of experiments to simultaneously shrink the weight of less significant factor, while looking for most informative settings to be tested. We have also developed simple sequential design strategies for efficient determination of subsequent experiments based on the information from previous experiments. To leverage the intrinsic geometry of factor settings in highly nonlinear spaces, we adapt the methodology to nonlinear and black-box problems using gradient directed kernels.

 

Research grants: This research line has so far resulted in $371,937 Young Investigator Award of Air force Office of Scientific Research (AFOSR-YIP) (FA9550-16-1-0171). It has also brought about a grant of $99,723 from Air Force Research Laboratory (FA8650-13-C-5800).

 

 

GP   

SLRD

Kriging

KBAR, Variant 1

KBAR, Variant 2

 

Research Thrust 3:  Improvement of Machine-to-Machine Reproducibility with Kernel-based Analysis: A Form 3+ SLA 3D Printing Study

Commercial 3D printers often face a high failure rate, with many even exceeding 20%. As 3D printing gains prominence in creating practical-use parts, ensuring reproducibility of desired dimensions becomes crucial. Numerous parameters in 3D printing impact part quality, making the quest for ideal reproducibility challenging and resource intensive. This becomes even more of a challenge when different printers are used.

 

This research project focuses on leveraging machine learning, specifically a kernelized approach within artificial intelligence, to enhance additive manufacturing operations, commonly known as 3D printing. Our project endeavors to enhance the reproducibility of printed parts by utilizing a combination of printer parameters and machine learning algorithms. This improvement is targeted not only within the same printer but also extends to achieving consistency across different printers. The current model employs kernelized machine learning, mapping input data to target output, with the ultimate goal of predicting optimal printer parameters for achieving the highest-quality parts.

 

 

image.png

 

Research Thrust 4:  AI-Driven Prediction, Monitoring, and Management of Unwanted Behavior in Patients with Autism
The research project is dedicated to leveraging artificial intelligence (AI) and machine learning techniques to address the challenges associated with Autism Spectrum Disorder (ASD). With statistics from the Centers for Disease Control and Prevention indicating that approximately 1 in 36 children are affected by ASD, there exists a pressing need to understand and effectively manage the behaviors that accompany this condition to enhance the well-being of individuals on the autism spectrum.

In this endeavor, advanced machine learning algorithms are being developed to analyze diverse data sources. These sources include behavioral patterns and physiological indicators, which play a crucial role in predicting and identifying patterns associated with unwanted behaviors in individuals with ASD. The primary goal is to facilitate the early detection of potential triggers that may lead to these behaviors.

Furthermore, the research is driven by the aim of creating personalized intervention strategies tailored to the unique characteristics of each individual. By doing so, the aspiration is to significantly improve the quality of life for individuals with autism while also providing invaluable support for their caregivers.

 

Research Thrust 5:  Airforce
The research project is dedicated to leveraging artificial intelligence (AI) and machine learning techniques to address the challenges associated with Autism Spectrum Disorder (ASD). With statistics from the Centers for Disease Control and Prevention indicating that approximately 1 in 36 children are affected by ASD, there exists a pressing need to understand and effectively manage the behaviors that accompany this condition to enhance the well-being of individuals on the autism spectrum.

In this endeavor, advanced machine learning algorithms are being developed to analyze diverse data sources. These sources include behavioral patterns and physiological indicators, which play a crucial role in predicting and identifying patterns associated with unwanted behaviors in individuals with ASD. The primary goal is to facilitate the early detection of potential triggers that may lead to these behaviors.

Furthermore, the research is driven by the aim of creating personalized intervention strategies tailored to the unique characteristics of each individual. By doing so, the aspiration is to significantly improve the quality of life for individuals with autism while also providing invaluable support for their caregivers.

 

Research Thrust 6:  Urban Heat Island Analysis and Mitigation through Digital Twin
The project aims to address the Urban Heat Island (UHI) effect in San Antonio by developing an advanced digital twin platform. This platform is designed to meticulously replicate the urban environment of the city, enhancing its resilience. By integrating real-time data from various sources such as weather stations, sensors, and satellite imagery, the digital twin will provide a dynamic representation of the city's urban fabric, climate patterns, and UHI hotspots. Key features of the platform include predictive modeling, scenario analysis, and decision support, all aimed at facilitating evidence-based interventions and policies to effectively mitigate the impact of UHI.

bottom of page