Document Type

Thesis

Degree

Master of Science

Major

Computer Science

Date of Defense

11-21-2024

Graduate Advisor

Sharlee Climer

Committee

Sharlee Climer

Mark Hauschild

Badri Adhikari

Abstract

Large-scale, high-dimensional data analyses can be computationally prohibitive due to combinatorial explosion of the search space for finding complex patterns; a viable alternative is network modeling for abstraction and quantifying intrinsic data associations. Prominent network analysis methods furnish frameworks for model synthesis and validation but rely on standard correlation measures impaired by semi-supervised biases, latent heterogeneity, and uneven discretization techniques. Here we investigate a holistic measure for encapsulating data heterogeneity for enhanced efficacy of revealing complex patterns through network analysis. Our unique correlation metric, K-medoids Utility for Duo Original Similarities (Kudos), exhaustively factors real-valued analyte data to compute a vector-based measure for useful network construction. We present compelling results for our application towards quantifying statistically significant, synchronized genetic interactions associated with late-onset Alzheimer’s disease. In contrast with established measures, we observed unprecedent pattern association odds ratios following independent validation, indicating our developed metric renders enhanced perceptivity and encapsulation of complex trait heterogeneity. We offer our thorough network composition and validation analyses for extension into related and unspecified domain applications.

Share

COinS