DSpace@RPI (Rensselaer Polytechnic Institute)
Not a member yet
6809 research outputs found
Sort by
LLM-Based Code Generation for Querying Temporal Tabular Financial Data
No full textWe examine the question of ``how well large language models (LLMs) can answer questions using temporal tabular financial data by generating code?''. Leveraging advanced language models, specifically GPT-4 and Llama 3, we aim to scrutinize and compare their abilities to generate coherent and effective code for Python, R, and SQL based on natural language prompts. We design an experiment to assess the performance of LLMs on natural language prompts on a large temporal financial dataset. We created a set of queries with hand-crafted R code answers. To investigate the strengths and weaknesses of LLMs, each query was created with different factors that characterize the financial meaning of the queries and their complexity. We demonstrate how to create specific zero-shot prompts to generate code to answer natural language queries about temporal financial tabular data. We develop specific system prompts for each language to ensure they correctly answer time-oriented questions. We execute this experiment on two LLMs (GPT-4 and Llama 3), assess if the outputs produced are executable and correct, and assess the efficiency of the produced code for Python, SQL, and R. We find that while LLMs have promising performance, their performance varies greatly across the languages, models, and experimental factors. GPT-4 performs best on Python (95.2\% correctness) but has significantly weaker performance on SQL (87.6\% correctness) and R (79.0\% correctness). Llama 3 is less successful at generating code overall, but it achieves its best results in R (71.4\% correctness). A multi-factor statistical analysis of the results with respect to the defined experimental factors provides further insights into the specific areas of challenge in code generation for each LLM. Our preliminary results on this modest benchmark demonstrate a framework for developing larger, comprehensive, unique benchmarks for both temporal financial tabular data and R code generation. While Python and SQL already have benchmarks, we are filling in the gaps for R. Powerful AI agents for text-to-code generation, as demonstrated in this work, provide a critical capability required for the next-generation AI-based natural language financial intelligence systems and chatbots, directly addressing the complex challenges presented by querying temporal tabular financial data
Development of light-responsive, fast-setting photocatalytic radical acrylic resins with enhanced biomechanical properties for the augmentation of osteoporotic vertebral compression fractures
No full textAugust 2024School of EngineeringThe management of osteoporotic vertebral compression fractures (VCFs) critically depends on the biomechanical, chemical, and physical properties of bone cement. Recent advancements in this field have involved the incorporation of bioactive molecules and adjustments in reactant-substrate ratios to enhance both the operative and mechanical characteristics of bone cement. Despite progress, commercially available cement still lacks the ability for surgeons to control setting properties in real-time, which is crucial to prevent cement extrusion and minimize air pocket formation during vertebral augmentation. Moreover, fast-setting cements often present a mismatch in mechanical properties with native vertebral bone, leading to poor stress distribution and potential adjacent vertebral fractures. This research identifies three controllable, fast-setting injectable photocatalytic radical acrylic resins that aim to mimic the mechanical properties of vertebral cortico-cancellous bone and maintain peak temperatures below thresholds likely to cause heat-induced tissue damage. Utilizing UV light to initiate the setting, this approach allows for precise control over the resin hardening time. The dissertation employs a needs-based translational framework to thoroughly assess the physical, chemical, and mechanical characteristics of the photocatalytic radical acrylic resins, identifying key factors to overcome current limitations in VCF augmentation. The photo-active cement utilized in this study is primarily composed of photoinitiators such as Camphorquinone (CQ), lithium phenyl-2,4,6-trimethylbenzoylphosphinate (lap), and 2,2-dimethoxy-2-phenylacetophenone (DMPA), with benzoyl peroxide (BPO) serving as an accelerator. The incorporation of BPO into photo-active bone cement significantly improved their setting times, although it did not enhance the bending strength or Young’s modulus compared to traditional acrylic formulations. Intriguingly, these modifications resulted in mechanical properties that closely align with those of healthy vertebral bone, a benchmark that traditional cements fail to meet. While these findings do not conform to the industry standards specified in iso 5833:2002, they raise compelling questions about the potential for developing biomimetic mechanical properties in acrylic bone cement. This discrepancy challenges the current industry norms and suggests a shift towards tailoring bone cement properties to more closely mimic natural bone, potentially leading to improved clinical outcomes in orthopedic and spinal surgeries.Ph
MortalityMinder: Visualization and AI Interpretations of Social Determinants of Premature Mortality in the United States
No full textMortalityMinder enables healthcare researchers, providers, payers, and policy makers to gain actionable insights into where and why premature mortality rates due to all causes, cancer, cardiovascular disease, and deaths of despair rose between 2000 and 2017 for adults aged 25–64. MortalityMinder is designed as an open-source web-based visualization tool that enables interactive analysis and exploration of social, economic, and geographic factors associated with mortality at the county level. We provide case studies to illustrate how MortalityMinder finds interesting relationships between health determinants and deaths of despair. We also demonstrate how GPT-4 can help translate statistical results from MortalityMinder into actionable insights to improve population health. When combined with MortalityMinder results, GPT-4 provides hypotheses on why socio-economic risk factors are associated with mortality, how they might be causal, and what actions could be taken related to the risk factors to improve outcomes with supporting citations. We find that GPT-4 provided plausible and insightful answers about the relationship between social determinants and mortality. Our work is a first step towards enabling public health stakeholders to automatically discover and visualize relationships between social determinants of health and mortality based on available data and explain and transform these into meaningful results using artificial intelligence.United Health Foundation; The Rensselaer IDE
An ontology-enabled approach for user-centered and knowledge-enabled explanations of ai systems
No full textAugust2024School of ScienceExplainability has been a well-studied problem in the Artificial Intelligence (AI) community through several AI ages, ranging from expert systems to the current deep learning era, to enable AI's safe and robust use. Through the ages, the unique nature of AI approaches and their applications have necessitated the explainability approaches to evolve as well. However, these multiple iterations of explaining AI decisions have all focused on helping humans better understand and analyze the results and workings of AI systems. In this thesis, we seek to further the user-centered explainability sub-field in AI by addressing several challenges around explanations in the current AI era, which is characterized by the availability of many machine learning (ML) explainers and neuro-symbolic AI approaches. Previous research in user-centered explainability has mainly focused on what needs to be explained and less so on implementations for them. Additionally, there are challenges to supporting explanations in a manner that humans can easily interpret due to the lack of a unified framework for different explanation types and methods to support domain knowledge from authoritative literature. We address the three challenges or research questions around user-centered explainability: How can we formally represent explanations with support for interacting AI systems (AI methods in applied ecosystem), additional data sources, and along different dimensions? How useful and feasible are such explanations for clinical settings? Is it feasible to combine explanations from multiple data modalities and AI methods? For the first research question, we design an Explanation Ontology (EO), a general-purpose semantic representation that can represent fifteen different literature-derived explanation types via their system-, interface- and user- related components. We demonstrate the utility of the EO in representing explanations across different use cases, supporting system designers to answer explanation-related questions via a set of competency questions, and categorizing explanations to be of supported explanation types within our ontology. For the second research question, we focus on key explanation dimension, that is, contextual explanation and conduct a case study on supporting contextual explanations from an authoritative knowledge source, clinical practice guidelines (CPGs). Here, we design a clinical question-answering (QA) system to address CPG questions to provide contextual explanations to help clinicians interpret risk prediction scores and their post hoc explanations in a comorbidity risk prediction setting. For the QA system, we leverage large language models (LLMs) and their clinical variants and implement knowledge augmentations to these models to improve semantic coherence of the answers. We evaluate both the feasibility and value of supporting these contextual explanations. For feasibility, we use quantitative metrics to report the performance of the QA system and do so across different model settings and data splits. To evaluate the value of the explanations, we report findings from showing the results of our QA approach to an expert panel of clinicians. Finally, for the last research question, we design a general-purpose and open-source framework, Metaexplainer, capable of providing natural-language explanations to a user question from the several explainer methods registered to generate explanations of a particular type. The Metaexplainer is a three-stage (Decompose, Delegate, and Synthesis) modular framework through which each stage produces intermediate outputs that the next stage ingests. In the Decompose stage, we input user questions and identify what explanation type can best address them and generate actionable machine interpretations; in the Delegate stage, we run explainer methods registered for the identified explanation type and pass filters if any, from the question and finally, in the synthesis stage we generate natural-language explanations along the explanation type template. For the Metaexplainer, we leverage LLMs, the EO, and explainer methods to generate user-centered explanations in response to user questions. We evaluate the Metaexplainer on open-source tabular datasets, but the framework can be applied to other modalities with code adaptations. Overall, through this thesis, we aim to design methods that can support knowledge-enabled explanations across different use cases, accounting for the methods in today's AI era that can generate the supporting components of these explanations and domain knowledge sources that can enhance them. We demonstrate the efficacy of our approach in two clinical use cases as case studies but design our methods to be applied to use cases outside of healthcare as well. By implementing approaches for knowledge-enabled explainability that leverage the strengths of symbolic and neural AI, we take a step towards user-centered explainability to help humans interpret and understand AI decisions from different perspectives.Ph
Modeling and simulation of fuel dispersal during the loss-of-coolant accident
No full textAugust 2024School of EngineeringAs the nuclear industry explores new fuel designs to accommodate increased burnup, studying fuel behavior during loss-of-coolant accidents (LOCA) is essential for ensuring the safe operation of light-water reactors. In the event of cladding rupture during a postulated LOCA in a pressurized water reactor, fuel particles, along with fission gases, can be expelled into the reactor core from the fractured fuel rod. This expulsion, known as fuel dispersal, poses thermal hydraulic and radiological challenges, as the transport and resulting mass distribution of fuel fragments within the reactor pressure vessel can affect long-term cooling. The initial fuel dispersal phenomenon is significantly influenced by the high-pressure ejection characteristics of the fuel fragments, the size and shape of the ruptured cladding, and the fuel rod depressurization history during LOCA transients. Depending on the location of the burst orifice relative to the quench front, the dispersal event represents an intricate three-phase flow and heat transfer phenomenon, where high-temperature fuel particles carried by the fission gases induce phase changes within the the narrow subchannels of the fuel assemblies. Given the unique nature of this flow problem, the current study develops a dedicated computational framework to predict the mass distribution and cooling of dispersing fuel particles, facilitating post-accident management of the fuel assemblies. Firstly, considering the scale of nuclear reactor applications, a continuum three-fluid model is proposed for simulating the transport of solids within the reactor core. With high-temperature fuel fragments within the liquid media, nucleation sites inducing phase changes are dispersed within the flow domain. Coupled with the fact that the transient dispersal event occurs on different time scales than other three-phase flow applications, this study derives a time-averaged three-fluid flow model without losing generality. The assumptions regarding the continuum treatment of the solid phase and the modeling of fuel dispersal behavior are incorporated to simplify the governing equations and derive applicable closure relations. The computational validation of the model was conducted using adiabatic experimental results obtained from ongoing research at Oregon State University, focusing on characterizing fuel dispersal behavior during simulated LOCA conditions. Settlement characteristics of the solids, quantified by the probability distribution of equivalent particles, closely matched the probability density functions reported in experimental studies.
To gain insights into the governing flow regime characteristics during a dispersal event, a discrete phase modeling framework was developed to track solid particles in a Lagrangian manner within the gas-liquid flow. A two-way coupling algorithm was proposed, where interfacial and buoyancy forces are updated for the particle phase based on their instantaneous positions within the computational domain. Simultaneously, the forcing terms due to particle loading within a computational cell are updated through a source term for the gas-liquid two-fluid flow. This modeling framework was validated using adiabatic experimental results from the literature, and the particle settlement characteristics were found to be in good agreement.
Lastly, the transport of fuel particles within a scaled 5 × 5 lattice of a pressurized-water reactor rod bundle geometry was modeled through a two-fluid Eulerian framework. The required boundary conditions were evaluated from the fuel performance code BISON in a postulated large-break LOCA scenario. The modeling framework considered solid fuel particles as granular matter, interacting with the gaseous dry steam phase and fission gases through the governing interfacial momentum exchange between the participating fluids. The simulation results provided the volume fraction of the solids obtained at the bottom surface of the enclosing tank geometry.
The key highlights of this study include a detailed derivation of a three-fluid flow modeling framework, a theoretical description of fuel dispersal behavior, and simplifying assumptions necessary to arrive at a set of governing equations and appropriate closure models for interpenetrating three-phase flow during a simulated fuel dispersal event. Additionally, the study demonstrates a three-fluid coupling algorithm for tracking inert point masses interacting with gas-solid flows. Finally, it discusses key factors affecting fuel dispersal behavior within reactor rod bundle geometry during a postulated LOCA event in a PWR.Ph
Stochastic kinetic theory applied to coarse-grained polymer model
No full textMay 2024School of EngineeringThe polymer fluid coupled with the flow could introduce inhomogeneities into the polymer concentration, leading to phase separation and other phenomena. Over the decade, many models have been established to understand polymer fluid behaviors. This thesis focuses on applying the stochastic kinetic theory approach to polymer systems to better understand the concentrated polymer system out of equilibrium. The stochastic kinetic theory has been used in particle studies, such as interactive particles and active matter, but it has never been applied to objects with internal degrees of freedom. This thesis adapts a coarse-grained bead-spring chain model representing polymers in systems. The polymers are modeled as Hookean dumbbells without polymer-polymer and hydrodynamic interactions in one-dimensional and two-dimensional systems with a simple shear flow. The stochastic moment equations are derived from the full field theory. The validation of the new stochastic field theory approach and the moment equations is compared to the explicit Langevin equation under conditions with analytical solutions. The time correlation function is used to quantify the stochastic behavior of the field theory. The moment equations are proven to have better accuracy at all ranges of polymer numbers without tracking the polymer configuration, while the stochastic field theory is only accurate at large numbers of polymer. Both stochastic models experience errors in the scale closer to a single mesh point due to their coarse-grained nature. In two-dimensional systems, the effect of the shear flow is examined, and the model behaviors in flow direction are quantified with Taylor dispersion analysis.Ph
Fluid role structure in teams: coordination and performance in complex cooperative tasks
No full textAugust 2024School of Humanities, Arts, and Social SciencesHuman behavior in a task is largely determined by its context, which include the characteristics of the task and its environment. With human teams, there is an additional level of contextual complexity arising from the properties of the team itself (size, composition, and degree of interdependence to name a few). This creates challenges for the generalizability of team research. Team coordination studies typically focus on teams in which members have highly specialized skills, and roles that complement these skills. Here, I take an exploratory approach to study changes in coordination strategies over time and their effect on performance, for teams with fluid role structures. Additionally, both individual and team level measures of behavior have been considered for the analyses, based on recent recommendations. Team performance data was collected from a cooperative cooking game across 8 one-hour sessions. In addition, 3 cognitive test sessions were also conducted. Results suggest that while planned role coordination strategies were beneficial to team performance towards the beginning of the study, adaptive strategies resulted in better performance after sufficient practice. Further, the average rate of interaction within a team was found to be positively correlated with performance, with significant performance effects associated with a small number of highly interactive or independent players. Finally, examination of cognitive correlates of team skill revealed that participants shifted from relying on visuospatial working memory earlier in the study to attentional control abilities towards the end, for team coordination. These results carry important implications for the design and training of essential service teams such as firefighter, army, paramedic, and search and rescue teams.Ph
Causal learning via interventions: estimation and design
No full textMay 2024School of EngineeringThe focus of this dissertation is leveraging the interventions in causal learning. Directed acyclic graphs (DAGs) have been used to compactly represent cause-effect relationships among random variables. Interventions refer to the changes in the causal mechanisms that govern the causal model and they are essential to causality in many aspects, such as causal structure learning, causal representation learning, and predicting the effects of an action in causal models. This dissertation broadly investigates the estimation and design of interventions in these aspects. We first consider the problem of estimating unknown intervention targets in causal models. We are motivated by two significant challenges. First, most of the interventional structure learning algorithms require the knowledge of the intervention targets. Secondly, there exist application domains in which the goal is to learn only the differences between two related networks. We take a direct approach and decouple the estimation of intervention targets from learning the causal structure in its entirety, which is the key to developing scalable algorithms. We focus on linear structural equation models (SEMs) with soft interventions and propose a framework that consists of a sequence of precision difference estimation steps. We study both causally sufficient and insufficient models, and design efficient algorithms with consistency guarantees. Our method can also be used in conjunction with existing observational learning algorithms to refine their results when interventional data becomes available. Finally, we establish sample complexity guarantees of our algorithms. Next, we consider using interventions to learn latent causal models and representations. Causal variables are not always observed directly but can be latent, only observed through an unknown transformation function. Causal representation learning (CRL) aims to invert the data generation process of high-dimensional observations and recover the underlying causal variables and their relationships. We proposed a novel framework (Score-based CRL) that leverages interventional data and properties of the score functions (gradient of log-likelihood) to identify latent causal variables and their relationships. Our key idea is that true inverse transform can be recovered by minimizing the number of changes in the score functions of latent variables across interventional environments. Our work establishes the first identifiability results for nonparametric causal models under a linear transform, and the first provably correct algorithm for the setting of general nonparametric transformations while also removing faithfulness assumptions on causal models. The third focus of this dissertation is designing an optimal sequence of interventions in a given causal graph to maximize the expected value of a reward node. This is, naturally, posed as a causal bandit problem. The majority of the previous works on causal bandits assume that interventional distributions are at least partially known. Acquiring such knowledge becomes prohibitive even in moderate-sized graphs when the set of possible interventions grows exponentially with the number of nodes. We address this issue on linear SEMs as follows: We avoid directly estimating exponentially many reward distributions and instead estimate the parameters that fully specify the SEMs to compute the rewards. Under boundedness assumptions on noise and the parameter space, we establish upper bounds on the regrets of proposed algorithms and a minimax lower bound for the problem. Remarkably, the achievable and lower bounds conform in their scaling behavior with respect to the horizon and graph parameters and are not affected by the cardinality of the intervention space. Finally, this dissertation investigates causal discovery of a mixture of DAGs. In some complex systems, such as time-varying causal systems, specifying all causal relationships with a single DAG is not viable. We aim to address this problem and formalize a framework for recovering the causal graph representing a mixture of DAGs. Specifically, we introduce the notion of emergent edges, which are the causal variables that are nonadjacent in individual models but become connected in the mixture. We establish the necessary and sufficient conditions for emergence of these spurious connections for a mixture of DAGs and propose an algorithm to infer the learnable causal relations. Furthermore, we analyze the necessary and sufficient interventions on a mixture of DAGs to identify these spurious connections.Ph
Bergeron: combating adversarial attacks by emulating a conscience
No full textMay 2024School of ScienceArtificial Intelligence alignment is the practice of encouraging an AI to behave in a manner that is compatible with human values and expectations. Research into this area has grown considerably since the recent introduction of increasingly capable Large Language Models (LLMs). The most effective contemporary methods of alignment are primarily weight-based: modifying the internal weights of a model to better align its behavior with human preferences. An optimal alignment process results in an AI model that is maximally helpful to its user while generating minimally harmful responses. Unfortunately, modern methods of alignment still fail to fully prevent harmful responses when faced with effective adversarial attacks. These deliberate attacks can trick seemingly aligned models into giving manufacturing instructions for dangerous materials, inciting violence, or recommending other immoral acts. To help mitigate this issue, I introduce Bergeron: a framework designed to improve the robustness of LLMs against attacks without any additional parameter fine-tuning. Bergeron is organized into two tiers; with a secondary LLM emulating the conscience of a protected, primary LLM. This framework better safeguards the primary model against incoming attacks while monitoring its output for any harmful content. Empirical analysis shows that, by using Bergeron to complement models with existing alignment training, we can improve the robustness and safety of multiple, commonly used commercial and open-source LLMs. Additionally, I demonstrate that a carefully chosen secondary model can effectively protect even much larger primary LLMs with a relatively minimal impact on Bergeron's resource usage.M
Angled layer deposition for advanced part geometries with wire arc additive manufacturing
No full textAugust 2024School of EngineeringThis work examines methods to create variable height layers using Wire Arc Additive Manufacturing (WAAM), a metal additive manufacturing process popular for its high deposition rate. We will examine how the torch speed can be used to control the height of the bead along each layer of a thin-walled part. Using this technique we are able to manufacture a part with an overhang that would traditionally only be possible using support material or geometric decomposition techniques. We then examine how the width of these parts changes as a function of the torch speed and feed rate of the welder. Using this relationship, we attempt to control the height and width of a thin-walled part independently. We then amend existing overlapping bead models to create angled layers for parts with widths larger than a single bead.M