<div>Complex systems are often characterized by a large collection of components interacting in nontrivial ways. Self-organization among these individual components often leads to emergence of a macroscopic structure that is neither completely regular nor completely random. In order to understand what we observe at a macroscopic scale, conceptual, mathematical, and computational tools are required for modeling and analyzing these interactions. A principled approach to understand these complex systems (and the processes that give rise to them) is to formulate generative models and infer their parameters from given data that is typically stored in the form of networks (or graphs). The increasing availability of network data from a wide variety of sources, such as the Internet, online social networks, collaboration networks, biological networks, etc., has fueled the rapid development of network science. </div><div><br></div><div>A variety of generative models have been designed to synthesize networks having specific properties (such as power law degree distributions, small-worldness, etc.), but the structural richness of real-world network data calls for researchers to posit new models that are capable of keeping pace with the empirical observations about the topological properties of real networks. The mechanistic approach to modeling networks aims to identify putative mechanisms that can explain the dependence, diversity, and heterogeneity in the interactions responsible for creating the topology of an observed network. A successful mechanistic model can highlight the principles by which a network is organized and potentially uncover the mechanisms by which it grows and develops. While it is difficult to intuit appropriate mechanisms for network formation, machine learning and evolutionary algorithms can be used to automatically infer appropriate network generation mechanisms from the observed network structure.</div><div><br></div><div>Building on these philosophical foundations and a series of (not new) observations based on first principles, we extrapolate an action-based framework that creates a compact probabilistic model for synthesizing real-world networks. Our action-based perspective assumes that the generative process is composed of two main components: (1) a set of actions that expresses link formation potential using different strategies capturing the collective behavior of nodes, and (2) an algorithmic environment that provides opportunities for nodes to create links. Optimization and machine learning methods are used to learn an appropriate low-dimensional action-based representation for an observed network in the form of a row stochastic matrix, which can subsequently be used for simulating the system at various scales. We also show that in addition to being practically relevant, the proposed model is relatively exchangeable up to relabeling of the node-types. </div><div><br></div><div>Such a model can facilitate handling many of the challenges of understanding real data, including accounting for noise and missing values, and connecting theory with data by providing interpretable results. To demonstrate the practicality of the action-based model, we decided to utilize the model within domain-specific contexts. We used the model as a centralized approach for designing resilient supply chain networks while incorporating appropriate constraints, a rare feature of most network models. Similarly, a new variant of the action-based model was used for understanding the relationship between the structural organization of human brains and the cognitive ability of subjects. Finally, our analysis of the ability of state-of-the-art network models to replicate the expected topological variations in network populations highlighted the need for rethinking the way we evaluate the goodness-of-fit of new and existing network models, thus exposing significant gaps in the literature.</div>

No description available.

In this work we use a graph-theoretic representation of secondary RNA structure found in the database RAG: RNA-As-Graphs. We model the bonding of two RNA secondary structures to form a larger structure with a graph operation called merge. The resulting data from each tree merge operation is summarized and represented by a vector. We use these vectors as input values for a neural network and train the network to recognize a tree as RNA-like or not based on the merge data vector. The network correctly assigned a high probability of RNA-likeness to trees identified as RNA-like in the RAG database, and a low probability of RNA-likeness to those classified as not RNA-like in the RAG database. We then used the neural network to predict the RNA-likeness of all the trees of order 9. The use of a graph operation to theoretically describe the bonding of secondary RNA is novel.

Polymer coatings offer a wide range of benefits across various industries, playing a crucial role in product protection and extension of shelf life. However, formulating them can be a non-trivial task given the multitude of variables and factors involved in the production process, rendering it a complex, high-dimensional problem. To tackle this problem, machine learning (ML) has emerged as a promising tool, showing considerable potential in enhancing various polymer and chemistry-based applications, particularly those dealing with high dimensional complexities. Our research aims to develop a physics-guided ML approach to facilitate the formulations of polymer coatings. As the first step, this project focuses on finding machine-readable feature representation techniques most suitable for encoding formulation ingredients. Utilizing two polymer-informatics datasets, one encompassing a large set of 700,000 common homopolymers including epoxies and polyurethanes as coating base materials while the other a relatively small set of 1000 data points of epoxy-diluent formulations, four featurization schemes to represent polymer coating molecules were benchmarked. They include the molecular access system, the extended connectivity fingerprint, molecular graph-based chemical graph network, and graph convolutional network (MG-GCN) embeddings. These representation schemes were used with ensemble models to predict molecular properties including topological surface area and viscosity. The results show that the combination of MG-GCN and ensemble models such as the extreme boosting machine and random forest models achieved the best overall performance, with coefficient of determination (r2) values of 0.74 in topological surface area and 0.84 in viscosity, which compare favorably with existing techniques. These results lay the foundation for using ML with physical modeling to expedite the development of polymer coating formulations.

This thesis explores a methodology to quantify trust in wearable medical devices (WMD) by addressing two main challenges: identifying key factors influencing trust and developing a formal framework for precise trust quantification under uncertainty. The work empirically validates trust factors and uses a Bayesian network to quantify trust. The thesis further employs a data-driven approach to estimate Bayesian parameters, facilitating query-based inference and validating the trust model with real and synthetic datasets, culminating in a customizable parameterized trust evaluation prototype for WMD. / Advances in sensor and digital communication technologies have revolutionized the capabilities of wearable medical device (WMD) to monitor patients’ health remotely, raising growing concerns about trust in these devices. There is a need to quantify trust in WMD for their continued acceptance and adoption by different users. Quantifying trust in WMD poses two significant challenges due to their subjective and stochastic nature. The first challenge is identifying the factors that influence trust in WMD, and the second is developing a formal framework for precise quantification of trust while taking into account the uncertainty and variability of trust factors. This thesis proposes a methodology to quantify trust in WMD, addressing these challenges. In this thesis, first, we devise a method to empirically validate dominant factors that influence the trustworthiness of WMD from the perspective of device users. We identified the users’ awareness of trust factors reported in the literature and additional user concerns influencing their trust. These factors are stepping stones for defining the specifications and quantification of trust in WMD. Second, we develop a probabilistic graph using Bayesian network to quantify trust in WMD. Using the Bayesian network, the stochastic nature of trust is viewed in terms of probabilities as subjective degrees of belief by a set of random variables in the domain. We define each random variable in the network by the trust factors that are identified from the literature and validated by our empirical study. We construct the trust structure as an acyclic-directed graph to represent the relationship between the variables compactly and transparently. We set the inter-node relationships, using the goal refinement technique, by refining a high-level goal of trustworthiness to lower-level goals that can be objectively implemented as measurable factors. Third, to learn and estimate the parameters of the Bayesian network, we need access to the probabilities of all nodes, as assuming a uniform Gaussian distribution or using values based on expert opinions may not fully represent the complexities of the factors influencing trust. We propose a data-driven approach to generate priors and estimate Bayesian parameters, in which we use data collected from WMD for all the measurable factors (nodes) to generate priors. We use non-functional requirement engineering techniques to quantify the impacts between the node relationships in the Bayesian network. We design propagation rules to aggregate the quantified relationships within the nodes of the network. This approach facilitates the computation of conditional probability distributions and enables query-based inference on any node, including the high-level trust node, based on the given evidence. The results of this thesis are evaluated through several experimental validations. The factors influencing trust in WMD are empirically validated by an extensive survey of 187 potential users. The learnability, and generalizability of the proposed trust network are validated with a real dataset collected from three users of WMD in two conditions, performing predefined activities and performing regular daily activities. To extend the variability of conditions, we generated an extensive and representative synthetic dataset and validated the trust network accordingly. Finally, to test the practicality of our approach, we implemented a user-configurable, parameterized prototype that allows users of WMD to construct a customizable trust network and effectively compare the trustworthiness of different devices. The prototype enables the healthcare industry to adapt and adopt this method to evaluate the trustworthiness of WMD for their own specific use cases. / Thesis / Doctor of Philosophy (PhD) / In this thesis, two challenges in quantifying trust in wearable medical devices, are addressed. The first challenge is the identification of factors influencing trust which are inherently subjective and vary widely among users. To address this challenge, we conducted an extensive survey to identify and validate the trust factors. These factors are stepping stones for defining the specifications and quantifying trust in wearable medical devices. The second challenge is to develop a precise method for quantification of trust while taking into account the uncertainty and variability of trust factors. We constructed a Bayesian network, that captures the complexities of trust as probabilities of the trust factors (identified from the survey) and developed a data-driven approach to estimate the parameters of the Bayesian network to compute the measure of trust. The findings of this thesis are empirically and experimentally validated across multiple use cases, incorporating real and synthetic data, various testing conditions, and diverse Bayesian network configurations. Additionally, we developed a customizable, parameterized prototype that empowers users and healthcare providers to effectively assess and compare the trustworthiness of different wearable medical devices.