Defense Notices

Discrete passive components, including resistors, capacitors, inductors, and transformers, are fundamental to broadband 75-Ω radio-frequency (RF) systems, particularly in cable television (CATV) and hybrid fiber-coaxial (HFC) networks. These components are incorporated into power splitters, couplers, power inserters, and amplifiers to achieve desired frequency responses, impedance matching, and signal transmission characteristics. However, their electrical behavior becomes increasingly complex at higher frequencies due to parasitic resistance, inductance, capacitance, electromagnetic coupling, and distributed transmission-line effects. Consequently, accurate broadband characterization is essential for predicting component behavior and developing representative RF circuit models for simulation purposes.

A significant challenge arises when characterizing passive components with limited or unavailable manufacturer specifications, particularly when existing component inventories must be evaluated for broadband applications. Conventional measurement instruments, including digital multimeters, LCR meters, oscilloscopes, and function generators, generally lack the frequency range required for comprehensive RF characterization from 5 MHz to 3 GHz. Vector network analyzers (VNAs) enable complex scattering-parameter measurements across this frequency range; however, the measured response includes contributions from test fixtures, interconnections, and parasitic discontinuities that must be distinguished from the intrinsic response of the device under test (DUT).

This project investigates the development of a signal flow graph (SFG)-based de-embedding framework for broadband passive linear multiport RF networks. The methodology incorporates series-through and shunt-through DUT (CPWG) fixture configurations, conventional Open-Short and Short-Open de-embedding techniques, followed by lumped-element and distributed transmission-line representations of fixture parasitics. Scattering-parameter and ABCD-parameter analysis are employed to characterize fixture responses, while signal flow graph derivations provide a mathematical basis for analyzing interconnected multiport networks. Particular attention is given to fixture asymmetry, impedance discontinuities, and multiple signal propagation paths that complicate the extension of conventional two-port characterization methods to higher-order networks.

Experimental validation is conducted through the characterization of passive components with published impedance and quality-factor characteristics, enabling comparisons between extracted measurements, equivalent-circuit models, and manufacturer reference data. The methodology is subsequently applied to broadband power splitter networks used in CATV/HFC systems to investigate the accuracy and applicability of the proposed multiport de-embedding framework. The project aims to establish a systematic approach to broadband passive component characterization that integrates fixture modeling, network theory, and experimental measurement while reducing dependence on proprietary characterization data.

Electronic health records (EHRs) contain medical codes and clinical notes collected over multiple visits. This dissertation studies how to learn patient representations from these records through three related aspects problems:

The first aspect concerns relationships within a patient’s history. TPGT models changing dependencies among medical codes within and across visits, and ARCI separates temporal patterns associated with different medical intents. They account for interacting clinical events and concurrent treatment goals.

When patient histories are limited, these patterns are harder to learn. The second aspect is addressed by MetaDrug, which adapts to patients with sparse histories using their own records and those of similar patients, while filtering uncertain samples.

The third aspect concerns incorporating clinical notes and large language models into patient representations. RePrompT uses cohort-trained representations and recurrent soft prompts to preserve longitudinal context. RASPER trains a note summarizer through downstream prediction feedback. CERE reasons over notes conditioned on structured codes to represent the additional evidence they provide. These studies connect structured histories with clinical text to support prediction.

Experiments on the KUMC acute kidney injury cohort, eICU, MIMIC-III, and MIMIC-IV show improvements over the respective baselines in clinical prediction and medication recommendation.

Large Language Models (LLMs) are becoming increasingly integrated into important domains such as social media and politics. This makes it essential that we understand their susceptibility to superficial text variations. In this thesis, I investigated whether small stylistic perturbations in social media posts influence how an LLM perceives a post's political leanings. To do this, I systematically manipulated punctuation and capitalization across both political and apolitical social media posts. I then used VADER sentiment analysis to measure the changes in semantic intensity to quantify the text manipulation. Finally, I prompted each LLM to provide its perceived political leanings of each given post. I used linear regression to assess the VADER score correlation with the LLMs' political alignment ratings. While there was no statistically significant relation between perceived political extremity and VADER scores, category-level analysis revealed that the stylistic perturbations cause significant inconsistencies in model outputs. Furthermore, in posts with neutral semantic content, the stylistic perturbations introduce a right-leaning bias. These findings add to the growing body of literature on LLM instability under prompt perturbation and highlight risks associated with automated content analysis. 

Accurate protein characterization, precision therapeutics design, and privacy-preserving clinical modeling are important challenges in biomedical artificial intelligence (AI). Addressing these challenges requires computational frameworks that are not only accurate and scalable but also capable of supporting reliable decision-making. This dissertation develops efficient, scalable, and reliable computational frameworks across three areas: Computational characterization of proteins, precision therapeutics design, and privacy-aware clinical representation learning. The first part of this dissertation focuses on efficient protein characterization at both the sequence and residue levels. GPCR-SLM, a knowledge-distilled small protein language model, classifies G-protein-coupled receptors (GPCRs), a family targeted by approximately 35% of FDA-approved drugs. Across 86 families, GPCR-SLM achieves 99% accuracy, outperforming BLAST (86.4%) and HMMER (91%), with a 33.5-fold speedup over large protein language models. Extending characterization to the residue level, a protein-metal-ion binding site prediction framework evaluates five protein language models and incorporates positional encoding to capture long-range dependencies. Under 10-fold cross-validation, it achieves a Matthews correlation coefficient of 0.89, with precision, recall, and F1 scores exceeding 95% across six metal ions.

Building on these protein characterization methods, the second part advances precision therapeutics design through uncertainty quantification and generative learning for therapeutic antibodies. NbBayesLM combines protein language model embeddings and physicochemical descriptors in a Bayesian neural network to predict nanobody thermostability. Evaluated on 10,630 VHH sequences, it achieves a mean absolute error of 1.89C and an R2 of 0.67 while estimating epistemic and aleatoric uncertainty to support candidate prioritization. CIgFlow employs classifier-free guidance to generate de novo, antigen-conditioned antibody sequences without fixed scaffolds or paired structural data. Generated sequences achieve a median OASis humanness score of 0.90 and mean predicted local distance difference test (pLDDT) score of 85.0, demonstrating promising novelty and diversity. The third part addresses privacy-aware clinical modeling using electrocardiograms (ECGs), which encode diagnostic information and sensitive demographic attributes. A variational autoencoder with dual discriminators suppresses demographic information while preserving clinically relevant representations.

Collectively, this dissertation advances efficient protein representation learning, computational antibody design, and privacy-preserving clinical AI, contributing methodological foundations for trustworthy biomedical applications.