Research

Abstract: Diagnostic data for Alzheimer’s Disease (AD) comes in many modalities, including MRI scans, demographic details, and biomarker assays. Computational AD diagnosis has become popular for timely treatment. However, current methods typically focus on a single modality to determine diagnosis. Our research delves into enhancing AD diagnosis accuracy by optimally fusing multimodal data through statistics and machine learning. We investigated techniques like PCA, autoencoders, LASSO, and proposed a method of dimensionality reduction called a “supervised encoder.”

Recognition: Presented at the 2023 International ACM “Knowledge Discovery and Data Mining” Conference

Research Paper: https://www.frontiersin.org/articles/10.3389/frdem.2024.1332928/full?

Code: https://github.com/MT-GoCode/Alzheimers-Data-Fusion

Abstract: My research focuses on developing novel shallow CNN architectures for plant disease diagnosis suitable for mobile devices, yet maintaining high accuracy. I compared my proposed CNNs against larger traditional models such as ResNet50, AlexNet, VGG16, and VGG19, all of which I implemented.

Recognition: 3rd Place Science Fair, 3rd Place in Congressional App Challenge

Research Paper: https://emerginginvestigators.org/articles/22-171

Code: https://github.com/MT-GoCode/ML-Computer-Vision-Demeter-Mobile-App

Abstract: My research focuses on developing novel shallow CNN architectures for plant disease diagnosis suitable for mobile devices, yet maintaining high accuracy. I compared my proposed CNNs against larger traditional models such as ResNet50, AlexNet, VGG16, and VGG19, all of which I implemented.

Abstract: Advanced to State Level in California State Science and Engineering Conference

Writeup: https://drive.google.com/file/d/1036-aFPXJg3INYDrzDApAz9U52_ZuArY/edit

Code: https://github.com/MT-GoCode/Reinforcement_Learning_Stock_Trading