Engineering Commons^™

Tutorial: Knowledge-Infused Artificial Intelligence For Mental Healthcare, Kaushik Roy

Publications

Artificial Intelligence (AI) systems for mental healthcare (MHCare) have been ever-growing after realizing the importance of early interventions for patients with chronic mental health (MH) conditions. Social media (SocMedia) emerged as the go-to platform for supporting patients seeking MHCare. The creation of peer-support groups without social stigma has resulted in patients transitioning from clinical settings to SocMedia supported interactions for quick help. Researchers started exploring SocMedia content in search of cues that showcase correlation or causation between different MH conditions to design better interventional strategies. User-level Classification-based AI systems were designed to leverage diverse SocMedia data from various MH conditions, …

K-Perm: Personalized Response Generation Using Dynamic Knowledge Retrieval And Persona-Adaptive Queries, Kanak Raj, Kaushik Roy, Vamshi Bonagiri, Priyanshul Govil, Krishnaprasad Thirunarayan, Raxit Goswami, Manas Gaur

Publications

Personalizing conversational agents can enhance the quality of conversations and increase user engagement. However, they often lack external knowledge to tend to a user’s persona appropriately. This is particularly crucial for practical applications like mental health support, nutrition planning, culturally sensitive conversations, or reducing toxic behavior in conversational agents. To enhance the relevance and comprehensiveness of personalized responses, we propose using a two-step approach that involves (1) selectively integrating user personas and (2) contextualizing the response with supplementing information from a background knowledge source. We develop K-PERM (Knowledge-guided PErsonalization with Reward Modulation), a dynamic conversational agent that combines these elements. …

Exploring Alternative Approaches To Language Modeling For Learning From Data And Knowledge, Yuxin Zi, Kaushik Roy, Vignesh Narayanan, Amit Sheth

Publications

Despite their wide applications to language understanding tasks, large language models (LLMs) still face challenges such as hallucinations - the occasional fabrication of information, and alignment issues - the lack of associations with human-curated world models (e.g., intuitive physics or common-sense knowledge). Additionally, the black-box nature of LLMs makes it highly challenging to train them meaningfully in order to achieve a desired behavior. Specifically, the attempt to adjust LLMs’ concept embedding spaces can be highly intractable, which involves analyzing the implicit impact on LLMs’ numerous parameters and the resulting inductive biases. This paper proposes a novel architecture that wraps powerful …

Causal Event Graph-Guided Language-Based Spatiotemporal Question Answering, Kaushik Roy, Alessandro Oltramari, Yuxin Zi, Chathurangi Shyalika, Vignesh Narayanan, Amit Sheth

Publications

Large Language Models have excelled at encoding and leveraging language patterns in large text-based corpora for various tasks, including spatiotemporal event-based question answering (QA). However, due to encoding a text-based projection of the world, they have also been shown to lack a fullbodied understanding of such events, e.g., a sense of intuitive physics, and cause-and-effect relationships among events. In this work, we propose using causal event graphs (CEGs) to enhance language understanding of spatiotemporal events in language models, using a novel approach that also provides proofs for the model’s capture of the CEGs. A CEG consists of events denoted by …

An Ontology Design Pattern For Representing Causality, Utkarshani Jaimini, Cory Henson, Amit Sheth

Publications

The causal pattern is a proposed ontology design pattern for representing the structure of causal relations in a knowledge graph. This pattern is grounded in the concepts defined and used by the CausalAI community i.e., Causal Bayesian Networks and do-calculus. Specifically, the pattern models three primary concepts: (1) causal relations, (2) causal event roles, and (3) causal effect weights. Two use cases involving a sprinkler system and asthma patients are provided along with their relevant competency questions.

Reducing Brain Kynurenic Acid Synthesis Precludes Kynurenine-Induced Sleep Disturbances, Katherine M. Rentschler, Snezana Milosavljevic, Annalisa M. Baratta, Courtney J. Wright, Maria V. Piroli, Zachary Tentor, Homayoun Valafar, Christian O'Reilly, Ana Pocivavsek

Publications

Patients with neurocognitive disorders often battle sleep disturbances. Kynurenic acid is a tryptophan metabolite of the kynurenine pathway implicated in the pathology of these illnesses. Modest increases in kynurenic acid, an antagonist at glutamatergic and cholinergic receptors, result in cognitive impairments and sleep dysfunction. We explored the hypothesis that inhibition of the kynurenic acid synthesising enzyme, kynurenine aminotransferase II, may alleviate sleep disturbances. At the start of the light phase, adult male and female Wistar rats received systemic injections of either: (i) vehicle; (ii) kynurenine (100 mg kg−1; i.p.); (iii) the kynurenine aminotransferase II inhibitor, PF-04859989 (30 mg kg−1; s.c.); …

Ki-Cook: Clustering Multimodal Cooking Representations Through Knowledge-Infused Learning, Revathy Venkataramanan, Swati Padhee, Saini Rohan Rao, Ronak Kaoshik, Anirudh Sundara Rajan, Amit Sheth

Publications

Cross-modal recipe retrieval has gained prominence due to its ability to retrieve a text representation given an image representation and vice versa. Clustering these recipe representations based on similarity is essential to retrieve relevant information about unknown food images. Existing studies cluster similar recipe representations in the latent space based on class names. Due to inter-class similarity and intraclass variation, associating a recipe with a class name does not provide sufficient knowledge about recipes to determine similarity. However, recipe title, ingredients, and cooking actions provide detailed knowledge about recipes and are a better determinant of similar recipes. In this study, …

Ecg Recordings As Predictors Of Very Early Autism Likelihood: A Machine Learning Approach, Deepa Tilwani, Jessica Bradshaw, Amit Sheth, Christian O'Reilly

Publications

In recent years, there has been a rise in the prevalence of autism spectrum disorder (ASD). The diagnosis of ASD requires behavioral observation and standardized testing completed by highly trained experts. Early intervention for ASD can begin as early as 1–2 years of age, but ASD diagnoses are not typically made until ages 2–5 years, thus delaying the start of intervention. There is an urgent need for non-invasive biomarkers to detect ASD in infancy. While previous research using physiological recordings has focused on brain-based biomarkers of ASD, this study investigated the potential of electrocardiogram (ECG) recordings as an ASD biomarker …

Analysis, Measurement, And Modeling Of Millimeter Wave Channels For Aviation Applications, Zeenat Afroze

Theses and Dissertations

Millimeter wave (mmWave) communication systems can employ a large amount of spectrum, and can consequently offer large data rates, e.g., multi-Gigabits-per-second. This technology can be used in many sectors: aviation, vehicles, public transportation, robotics, autonomous factories, etc. Yet mmWave communication systems suffer from some propagation challenges, including large free space path loss (PL), large penetration loss, and large diffraction loss. Hence, it is vital to quantify these and other channel effects to ensure link reliability. Most mmWave systems will employ directional antennas to enable acceptable link distances. In many settings this will require directional receiver antennas to rotate in azimuth …

Magnetic Softness Tuned Superparamagnetic Nanoparticles For Highly Efficient Cancer Theranostics, Jie Wang

Theses and Dissertations

Magnetic resonance imaging (MRI)-guided magnetic nanofluid hyperthermia (MNFH) using iron oxide based superparamagnetic nanoparticles (SPNPs) has recently attracted considerable attention as a treatment modality for cancer theranostics, because MRI-guided MNFH can allow for diagnosis, therapeutics, and prognosis simultaneously using the same administrated magnetic nanofluid agent. However, several primary limiting factors: (1) insufficient AC magnetic heating induction (specific loss power/intrinsic loss power, SLP/ILP) at the biologically safe and physically tolerable range of AC magnetic field (H_AC,safe: f_appl × H_appl < 3.0 ~ 5.0×10⁹ A·m^-1·s^-1), (2) low _r2- relaxivity directly related to the low resolution of …

Model-Driven Analysis Of Ecg Using Reinforcement Learning, Christian O'Reilly, Sai Durga Rithvik Oruganti, Deepa Tilwani, Jessica Bradshaw

Publications

Modeling is essential to better understand the generative mechanisms responsible for experimental observations gathered from complex systems. In this work, we are using such an approach to analyze the electrocardiogram (ECG). We present a systematic framework to decompose ECG signals into sums of overlapping lognormal components. We use reinforcement learning to train a deep neural network to estimate the modeling parameters from an ECG recorded in babies from 1 to 24 months of age. We demonstrate this model-driven approach by showing how the extracted parameters vary with age. From the 751,510 PQRST complexes modeled, 82.7% provided a signal-to-noise ratio that …

Novel Structures And Thin Film Techniques For Reconfigurable Rf Technologies With Improved Signal Integrity, Jinqun Ge

Theses and Dissertations

Over the past several decades, wireless communications have been tremendously developed with multiple frequency bands, standards, and functions to provide more convenient communication services. Nowadays, commercial wireless devices, especially customer premises equipment, are required for smaller size, higher integration, and higher data rates to meet the ever-increasing demands of communications. The transceiver front ends inside these devices need to provide high-speed and high-quality communication services without a substantial increase in cost and size. Hence, the primary objective of this research is to develop a few miniaturized, multifunctional, and reconfigurable RF technologies and to provide several signal integrity and heat dissipation …

Physics-Based And Behavioral Models For Fuel Cells, Charles Chima Anyim

Theses and Dissertations

In recent times there has been renewed interest in fuel cells for electricity production, especially for transportation applications. Fuel cells are attractive because the electricity production process from chemical energy does not cause pollution, and no carbon emissions occur. In this work, we will focus on Proton Exchange Membrane (PEM) fuel cells that use hydrogen and oxygen to generate electricity, producing only electricity, heat and water. If hydrogen can be produced in a green, non-polluting way, fuel cells can truly be an environmentally benign energy source.

In this work, we will look at the modeling of a fuel cell system. …

High-Performance Wide Bandgap Semiconductor Power Modules Enabled By Advanced Two-Phase Mini-Channel Cooling, Bo Tian

Theses and Dissertations

There is a widespread need for high performance wide bandgap power modules in both commercial and military applications. However, given the rapid advancements of wide bandgap power module technology, conventional cooling solutions have not kept up and do not provide the thermal management performance needed for high power density. Based on the two-phase cooling approach, two-phase microchannels operating at low fluid flow rates with low pressure drops have huge potential in enabling higher power density applications. Several studies have illustrated the potential great advantages of two-phase cooling compared to single-phase cooling in terms of maximum device temperature, spatial thermal distribution …

Quantized State Simulation Of Electrical Power Systems, Joseph Micah Hood

Theses and Dissertations

An alternative is proposed to the current state-of-the-art simulation methods for the transient simulation of electrical power systems. The proposed method combines the Latency Insertion Method (LIM), the Quantized Discrete Event Specification (QDEVS), and the Quantized State System (QSS) method of integration. Using LIM, the power system state equations are decoupled in a way that allows the formulation of the system into a QDEVS-compliant model, which can then be directly solved with various QSS integration techniques. This combination of methods is called the Quantized DEVS-LIM method, or simply, QDL. A key feature of QDL is the asynchronous updates of all …

Demo Alleviate: Demonstrating Artificial Intelligence Enabled Virtual Assistance For Telehealth: The Mental Health Case, Kaushik Roy, Vedant Khandelwal, Raxit Goswami, Nathan Dolbir, Jinendra Malekar, Amit Sheth

Publications

After the pandemic, artificial intelligence (AI) powered support for mental health care has become increasingly important. The breadth and complexity of significant challenges required to provide adequate care involve: (a) Personalized patient understanding, (b) Safety-constrained and medically validated chatbot patient interactions, and (c) Support for continued feedback-based refinements in design using chatbot-patient interactions. We propose Alleviate, a chatbot designed to assist patients suffering from mental health challenges with personalized care and assist clinicians with understanding their patients better. Alleviate draws from an array of publicly available clinically valid mental-health texts and databases, allowing Alleviate to make medically sound and informed …

Tutorial - Shodhguru Labs: Knowledge-Infused Artificial Intelligence For Mental Healthcare, Kaushik Roy

Publications

Artificial Intelligence (AI) systems for mental healthcare (MHCare) have been ever-growing after realizing the importance of early interventions for patients with chronic mental health (MH) conditions. Social media (SocMedia) emerged as the go-to platform for supporting patients seeking MHCare. The creation of peer-support groups without social stigma has resulted in patients transitioning from clinical settings to SocMedia supported interactions for quick help. Researchers started exploring SocMedia content in search of cues that showcase correlation or causation between different MH conditions to design better interventional strategies. User-level Classification-based AI systems were designed to leverage diverse SocMedia data from various MH conditions, …

Cook-Gen: Robust Generative Modeling Of Cooking Actions From Recipes, Revathy Venkataramanan, Kaushik Roy, Kanak Ray, Renjith Prasad, Yuxin Zi, Vignesh Narayanan, Amit Sheth

Publications

As people become more aware of their food choices, food computation models have become increasingly popular in assisting people in maintaining healthy eating habits. For example, food recommendation systems analyze recipe instructions to assess nutritional contents and provide recipe recommendations. The recent and remarkable successes of generative AI methods, such as auto-regressive large language models, can lead to robust methods for a more comprehensive understanding of recipes for healthy food recommendations beyond surface-level nutrition content assessments. In this study, we explore the use of generative AI methods to extend current food computation models, primarily involving the analysis of nutrition and …

Proknow: Process Knowledge For Safety Constrained And Explainable Question Generation For Mental Health Diagnostic Assistance, Kaushik Roy, Manas Gaur, Misagh Soltani, Vipula Rawte, Ashwin Kalyan, Amit Sheth

Publications

Current Virtual Mental Health Assistants (VMHAs) provide counseling and suggestive care. They refrain from patient diagnostic assistance because of a lack of training on safety-constrained and specialized clinical process knowledge (Pro-Know). In this work, we define ProKnow as an ordered set of information that maps to evidence-based guidelines or categories of conceptual understanding to experts in a domain. We also introduce a new dataset of diagnostic conversations guided by safety constraints and ProKnow that healthcare professionals use (ProKnow-data). We develop a method for natural language question generation (NLG) that collects diagnostic information from the patient interactively (ProKnow-algo). We demonstrate the …

Ierl: Interpretable Ensemble Representation Learning - Combining Crowdsourced Knowledge And Distributed Semantic Representations, Yuxin Zi, Kaushik Roy, Vignesh Narayanan, Manas Gaur, Amit Sheth

Publications

Large Language Models (LLMs) encode meanings of words in the form of distributed semantics. Distributed semantics capture common statistical patterns among language tokens (words, phrases, and sentences) from large amounts of data. LLMs perform exceedingly well across General Language Understanding Evaluation (GLUE) tasks designed to test a model’s understanding of the meanings of the input tokens. However, recent studies have shown that LLMs tend to generate unintended, inconsistent, or wrong texts as outputs when processing inputs that were seen rarely during training, or inputs that are associated with diverse contexts (e.g., well-known hallucination phenomenon in language generation tasks). Crowdsourced and …

Cooperative Deep Q -Learning Framework For Environments Providing Image Feedback, Krishnan Raghavan, Vignesh Narayanan, Sarangapani Jagannathan

Publications

In this article, we address two key challenges in deep reinforcement learning (DRL) setting, sample inefficiency, and slow learning, with a dual-neural network (NN)-driven learning approach. In the proposed approach, we use two deep NNs with independent initialization to robustly approximate the action-value function in the presence of image inputs. In particular, we develop a temporal difference (TD) error-driven learning (EDL) approach, where we introduce a set of linear transformations of the TD error to directly update the parameters of each layer in the deep NN. We demonstrate theoretically that the cost minimized by the EDL regime is an approximation …

Tutorial - Shodhguru Labs: Optimization And Hyperparameter Tuning For Neural Networks, Kaushik Roy

Publications

Neural networks have emerged as a powerful and versatile class of machine learning models, revolutionizing various fields with their ability to learn complex patterns and make accurate predictions. The performance of neural networks depends significantly on the appropriate choice of hyperparameters, which are critical factors governing their architecture, regularization, and optimization techniques. As the demand for high-performance neural networks grows across diverse applications, the need for efficient optimization and hyperparameter tuning methods becomes paramount. This paper presents a comprehensive exploration of optimization strategies and hyperparameter tuning techniques for neural networks. Neural networks have emerged as a powerful and versatile class …

The Troubling Emergence Of Hallucination In Large Language Models--An Extensive Definition, Quantification, And Prescriptive Remediations, Vipula Rawte, Swagata Chakraborty, Agnibh Pathak, Anubhav Sarkar, S.M Towhidul Islam Tonmoy, Aman Chadha, Amit Sheth, Amitava Das

Publications

The recent advancements in Large Language Models (LLMs) have garnered widespread acclaim for their remarkable emerging capabilities. However, the issue of hallucination has parallelly emerged as a by-product, posing significant concerns. While some recent endeavors have been made to identify and mitigate different types of hallucination, there has been a limited emphasis on the nuanced categorization of hallucination and associated mitigation methods. To address this gap, we offer a finegrained discourse on profiling hallucination based on its degree, orientation, and category, along with offering strategies for alleviation. As such, we define two overarching orientations of hallucination: (i) factual mirage (FM) …

Suitability Of Quantized Devs-Lim Methods For Simulation Of Power Systems, Navid Gholizadeh

Theses and Dissertations

This research presents an investigation of behavioral intricacies of the Quantized DEVS Latency Insertion Method (QDL) method and proposes resolutions to the previously unsolved and unexplained discrepancies between the reference solution and the QDL method. QDL method is a combination of two other methods namely the Latency Insertion Method (LIM) and Linear Implicit Quantized State (LIQSS)method. This technique is rigorously evaluated across a diverse array of systems and scenarios, with the aim of unearthing nuanced insights into its respective functionalities.

The research seeks to discern novel attributes of this method while gauging its comparative efficacy against conventional discrete-time methodologies, both …

Acm Web Conference 2023, Usha Lokala, Kaushik Roy, Utkarshani Jaimini, Amit Sheth

Publications

Improving the performance and explanations of ML algorithms is a priority for adoption by humans in the real world. In critical domains such as healthcare, such technology has significant potential to reduce the burden on humans and considerably reduce manual assessments by providing quality assistance at scale. In today’s data-driven world, artificial intelligence (AI) systems are still experiencing issues with bias, explainability, and human-like reasoning and interpretability. Causal AI is the technique that can reason and make human-like choices making it possible to go beyond narrow Machine learning-based techniques and can be integrated into human decision-making. It also offers intrinsic …

Event-Triggered Optimal Adaptive Control Of Partially Unknown Linear Continuous-Time Systems With State Delay, Rohollah Moghadam, Vignesh Narayanan, Sarangapani Jagannathan

Publications

This paper proposes an event-triggered optimal adaptive output feedback control design approach by utilizing integral reinforcement learning (IRL) for linear time-invariant systems with state delay and uncertain internal dynamics. In the proposed approach, the general optimal control problem is formulated into the game-theoretic framework by treating the event-triggering threshold and the optimal control policy as players. A cost function is defined and a value functional, which includes the delayed system output, is considered. First, by using the value functional and applying stationarity conditions using the Hamiltonian function, the output game delay algebraic Riccati equation (OGDARE) and optimal control policy are …

Tutorial: Knowledge-Infused Learning For Autonomous Driving (Kl4ad), Ruwan Wickramarachchi, Cory Henson, Sebastian Monka, Daria Stepanova, Amit Sheth

Publications

Autonomous Driving (AD) is considered as a testbed for tackling many hard AI problems. Despite the recent advancements in the field, AD is still far from achieving full autonomy due to core technical problems inherent in AD. The emerging field of neuro-symbolic AI and the methods for knowledge-infused learning are showing exciting ways of leveraging external knowledge within machine/deep learning solutions, with the potential benefits for interpretability, explainability, robustness, and transferability. In this tutorial, we will examine the use of knowledge-infused learning for three core state-of-the-art technical achievements within the AD domain. With a collaborative team from both academia and …

Tutorial: Neuro-Symbolic Ai For Mental Healthcare, Kaushik Roy, Usha Lokala, Manas Gaur, Amit Sheth

Publications

Artificial Intelligence (AI) systems for mental healthcare (MHCare) have been ever-growing after realizing the importance of early interventions for patients with chronic mental health (MH) conditions. Social media (SocMedia) emerged as the go-to platform for supporting patients seeking MHCare. The creation of peer-support groups without social stigma has resulted in patients transitioning from clinical settings to SocMedia supported interactions for quick help. Researchers started exploring SocMedia content in search of cues that showcase correlation or causation between different MH conditions to design better interventional strategies. User-level Classification-based AI systems were designed to leverage diverse SocMedia data from various MH conditions, …

Pulse Width Modulation-Based Voltage Balancing And Circulating Current Control For Modular Multilevel Converters, Md Multan Biswas

Theses and Dissertations

In the last decade, because of some key features and advantages of Modular Multilevel Converter (MMC), it is widely applied for various Medium Voltage (MV) and High Voltage (HV) power electronic interfacing applications. The structure of MMCs varies by application but in general, they are composed of multiple Sub-Modules (SMs) which each contains a floating capacitor. For the operation of MMCs, capacitor voltages are required to remain equally set. This is referred to as the balanced condition and requires a method for SM Capacitor Voltage Balancing (CVB). Connections of SMs for most of the common MMC configurations result in additional …

Metaversekg: Knowledge Graph For Engineering And Design Application In Industrial Metaverse, Utkarshani Jaimini, Tongtao Zhang, Georgia Olympia Brikis

Publications

While the term Metaverse was first coined by the author Neal Stephenson in 1992 in his science fiction novel “Snow Crash”, today the vision of an integrated virtual world is becoming a reality across different sectors. Applications in gaming and consumer products are gaining traction, industrial metaverse applications are, still in their early stages of development with one of the challenges being interoperability across various metaverse development platforms and existing software tools. In this work we propose the use of a knowledge graph based semantic data exchange layer, the Metaverse Knowledge Graph, to enable seamless transfer of information across platforms. …