Artificial Intelligence and Robotics Commons^™

Mosaic: A Prune-And-Assemble Approach For Efficient Model Pruning In Privacy-Preserving Deep Learning, Yifei Cai, Qiao Zhang, Rui Ning, Chunsheng Xin, Hongyi Wu

Computer Science Faculty Publications

To enable common users to capitalize on the power of deep learning, Machine Learning as a Service (MLaaS) has been proposed in the literature, which opens powerful deep learning models of service providers to the public. To protect the data privacy of end users, as well as the model privacy of the server, several state-of-the-art privacy-preserving MLaaS frameworks have also been proposed. Nevertheless, despite the exquisite design of these frameworks to enhance computation efficiency, the computational cost remains expensive for practical applications. To improve the computation efficiency of deep learning (DL) models, model pruning has been adopted as a strategic …

Identifying New Cancer Genes Based On The Integration Of Annotated Gene Sets Via Hypergraph Neural Networks, Chao Deng, Hong-Dong Li, Li-Shen Zhang, Yiwei Liu, Yaohang Li, Jianxin Wang

Computer Science Faculty Publications

Motivation

Identifying cancer genes remains a significant challenge in cancer genomics research. Annotated gene sets encode functional associations among multiple genes, and cancer genes have been shown to cluster in hallmark signaling pathways and biological processes. The knowledge of annotated gene sets is critical for discovering cancer genes but remains to be fully exploited.

Results

Here, we present the DIsease-Specific Hypergraph neural network (DISHyper), a hypergraph-based computational method that integrates the knowledge from multiple types of annotated gene sets to predict cancer genes. First, our benchmark results demonstrate that DISHyper outperforms the existing state-of-the-art methods and highlight the advantages of …

Can Large Language Models Discern Evidence For Scientific Hypotheses? Case Studies In The Social Sciences, Sai Koneru, Jian Wu, Sarah Rajtmajer

Computer Science Faculty Publications

Hypothesis formulation and testing are central to empirical research. A strong hypothesis is a best guess based on existing evidence and informed by a comprehensive view of relevant literature. However, with exponential increase in the number of scientific articles published annually, manual aggregation and synthesis of evidence related to a given hypothesis is a challenge. Our work explores the ability of current large language models (LLMs) to discern evidence in support or refute of specific hypotheses based on the text of scientific abstracts. We share a novel dataset for the task of scientific hypothesis evidencing using community-driven annotations of studies …

Charged Track Reconstruction With Artificial Intelligence For Clas12, Gagik Gavalian, Polykarpos Thomadakis, Angelos Angelopoulos, Nikos Chrisochoides

Computer Science Faculty Publications

In this paper, we present the results of charged particle track reconstruction in CLAS12 using artificial intelligence. In our approach, we use neural networks working together to identify tracks based on the raw signals in the Drift Chambers. A Convolutional Auto-Encoder is used to de-noise raw data by removing the hits that do not satisfy the patterns for tracks, and second Multi-Layer Perceptron is used to identify tracks from combinations of clusters in the drift chambers. Our method increases the tracking efficiency by 50% for multi-particle final states already conducted experiments. The de-noising results indicate that future experiments can run …

Short: Can Citations Tell Us About A Paper's Reproducibility? A Case Study Of Machine Learning Papers, Rochana R. Obadage, Sarah M. Rajtmajer, Jian Wu

Computer Science Faculty Publications

The iterative character of work in machine learning (ML) and artificial intelligence (AI) and reliance on comparisons against benchmark datasets emphasize the importance of reproducibility in that literature. Yet, resource constraints and inadequate documentation can make running replications particularly challenging. Our work explores the potential of using downstream citation contexts as a signal of reproducibility. We introduce a sentiment analysis framework applied to citation contexts from papers involved in Machine Learning Reproducibility Challenges in order to interpret the positive or negative outcomes of reproduction attempts. Our contributions include training classifiers for reproducibility-related contexts and sentiment analysis, and exploring correlations between …

Developing A Framework For Personalized Video-Based Quantum Information Science Education, Nikos Chrisochoides, Norou Diawara, Michail Giannakos

Computer Science Faculty Publications

This is a white paper on Workforce Development for Quantum Information Sciences (QIS) led by the Center for Real-Time Computing at Old Dominion University (ODU). We plan to investigate the potential of video lectures in supporting QIS. Specifically, we focus on following four objectives: (a) design a two-course series for both Master-level and PhD students; b) an upgrade of Experimental Lecture System (ELeSy) to test new, innovative, and transformative approaches for inclusive QIS education; c) design and implementation of a mixed-method systematic empirical study on the effects of video learning styles (in-person flipped classroom and voluntary video use) on graduate …

Bayesian Neural Netwok Variational Autoencoder Inverse Mapper (Bnn-Vaim) And Its Application In Compton Form Factors Extraction, Md Fayaz Bin Hossen, Tareq Alghamdi, Manal Almaeen, Yaohang Li

Computer Science Faculty Publications

We extend the Variational Autoencoder Inverse Mapper (VAIM) framework for the inverse problem of extracting Compton Form Factors (CFFs) from deeply virtual exclusive reactions, such as the unpolarized Deeply virtual exclusive scattering (DVCS) cross section. VAIM is an end-to-end deep learning framework to address the solution ambiguity issue in ill-posed inverse problems, which comprises of a forward mapper and a backward mapper to simulate the forward and inverse processes, respectively. In particular, we incorporate Bayesian Neural Network (BNN) into the VAIM architecture (BNN-VAIM) for uncertainty quantification. By sampling the weights and biases distributions of the BNN in the backward mapper …

Enhancing Heart Disease Prediction With Reinforcement Learning And Data Augmentation, Gayathri R., Sangeetha S. K. B., Sandeep Kumar Mathivanan, Hariharan Rajadurai, Benjula Anbu Malar Mb, Saurav Mallik, Hong Qin

Computer Science Faculty Publications

The study presents a novel method to improve the prediction accuracy of cardiac disease by combining data augmentation techniques with reinforcement learning. The complex nature of cardiac data frequently presents challenges for traditional machine learning models, which results in subpar performance. In response, our fusion methodology improves predictive capabilities by augmenting data and utilizing reinforcement learning's skill at sequential decision-making. Our method predicts cardiac disease with an astounding 94 % accuracy rate, which is an outstanding result. This significant improvement outperforms existing techniques and shows a deeper comprehension of intricate data relationships. The amalgamation of reinforcement learning and data augmentation …

Enhanced Skin Cancer Diagnosis Through Grid Search Algorithm-Optimized Deep Learning Models For Skin Lesion Analysis, Rudresh Pillai, Neha Sharma, Sheifali Gupta, Deepali Gupta, Sapna Juneja, Saurav Malik, Hong Qin, Mohammed S. Alqahtani, Amel Ksibi

Computer Science Faculty Publications

Skin cancer is a widespread and perilous disease that necessitates prompt and precise detection for successful treatment. This research introduces a thorough method for identifying skin lesions by utilizing sophisticated deep learning (DL) techniques. The study utilizes three convolutional neural networks (CNNs)-CNN1, CNN2, and CNN3-each assigned to a distinct categorization job. Task 1 involves binary classification to determine whether skin lesions are present or absent. Task 2 involves distinguishing between benign and malignant lesions. Task 3 involves multiclass classification of skin lesion images to identify the precise type of skin lesion from a set of seven categories. The most optimal …

Exacfs - A Cil Method To Mitigate Catastrophic Forgetting, S. Balasubramanian, Sai Subramaniam M., Sai Sriram Talasu, Manepalli Pranav Phanindra Sai, Yedu P. Krishna, Darshan Gera, Ravi Mukkamala

Computer Science Faculty Publications

Deep neural networks (DNNs) excel at learning from static datasets but struggle with continual learning, where data arrives sequentially. Catastrophic forgetting, the phenomenon of forgetting previously learned knowledge, is a primary challenge. This paper introduces EXponentially Averaged Class-wise Feature Significance (EXACFS) to mitigate this issue in the class incremental learning (CIL) setting. By estimating the significance of model features for each learned class using loss gradients, gradually aging the significance through the incremental tasks and preserving the significant features through a distillation loss, EXACFS effectively balances remembering old knowledge (stability) and learning new knowledge (plasticity). Extensive experiments on CIFAR-100 and …

Dilf: Differentiable Rendering-Based Multi-View Image-Language Fusion For Zero-Shot 3d Shape Understanding, Xin Ning, Zaiyang Yu, Lusi Li, Weijun Li, Prayag Tiwari

Computer Science Faculty Publications

Zero-shot 3D shape understanding aims to recognize “unseen” 3D categories that are not present in training data. Recently, Contrastive Language–Image Pre-training (CLIP) has shown promising open-world performance in zero-shot 3D shape understanding tasks by information fusion among language and 3D modality. It first renders 3D objects into multiple 2D image views and then learns to understand the semantic relationships between the textual descriptions and images, enabling the model to generalize to new and unseen categories. However, existing studies in zero-shot 3D shape understanding rely on predefined rendering parameters, resulting in repetitive, redundant, and low-quality views. This limitation hinders the model’s …

A Survey On Few-Shot Class-Incremental Learning, Songsong Tian, Lusi Li, Weijun Li, Hang Ran, Xin Ning, Prayag Tiwari

Computer Science Faculty Publications

Large deep learning models are impressive, but they struggle when real-time data is not available. Few-shot class-incremental learning (FSCIL) poses a significant challenge for deep neural networks to learn new tasks from just a few labeled samples without forgetting the previously learned ones. This setup can easily leads to catastrophic forgetting and overfitting problems, severely affecting model performance. Studying FSCIL helps overcome deep learning model limitations on data volume and acquisition time, while improving practicality and adaptability of machine learning models. This paper provides a comprehensive survey on FSCIL. Unlike previous surveys, we aim to synthesize few-shot learning and incremental …

Enabling Ai And Robotic Coaches For Physical Rehabilitation Therapy: Iterative Design And Evaluation With Therapists And Post-Stroke Survivors, Min Hun Lee, Daniel Siewiorek, Asim Smailagic, Alexandre Bernardino, Sergi Bermúdez I Badia

Research Collection School Of Computing and Information Systems

Artificial intelligence (AI) and robotic coaches promise the improved engagement of patients on rehabilitation exercises through social interaction. While previous work explored the potential of automatically monitoring exercises for AI and robotic coaches, the deployment of these systems remains a challenge. Previous work described the lack of involving stakeholders to design such functionalities as one of the major causes. In this paper, we present our efforts on eliciting the detailed design specifications on how AI and robotic coaches could interact with and guide patient’s exercises in an effective and acceptable way with four therapists and five post-stroke survivors. Through iterative …

Affinity Uncertainty-Based Hard Negative Mining In Graph Contrastive Learning, Chaoxi Niu, Guansong Pang, Ling Chen

Research Collection School Of Computing and Information Systems

Hard negative mining has shown effective in enhancing self-supervised contrastive learning (CL) on diverse data types, including graph CL (GCL). The existing hardness-aware CL methods typically treat negative instances that are most similar to the anchor instance as hard negatives, which helps improve the CL performance, especially on image data. However, this approach often fails to identify the hard negatives but leads to many false negatives on graph data. This is mainly due to that the learned graph representations are not sufficiently discriminative due to oversmooth representations and/or non-independent and identically distributed (non-i.i.d.) issues in graph data. To tackle this …

Continual Learning, Fast And Slow, Quang Anh Pham, Chenghao Liu, Steven C. H. Hoi

Research Collection School Of Computing and Information Systems

According to the Complementary Learning Systems (CLS) theory (McClelland et al. 1995) in neuroscience, humans do effective continual learning through two complementary systems: a fast learning system centered on the hippocampus for rapid learning of the specifics, individual experiences; and a slow learning system located in the neocortex for the gradual acquisition of structured knowledge about the environment. Motivated by this theory, we propose DualNets (for Dual Networks), a general continual learning framework comprising a fast learning system for supervised learning of pattern-separated representation from specific tasks and a slow learning system for representation learning of task-agnostic general representation via …

Conversational Localization: Indoor Human Localization Through Intelligent Conversation, Sheshadri Smitha, Kotaro Hara

Research Collection School Of Computing and Information Systems

We propose a novel sensorless approach to indoor localization by leveraging natural language conversations with users, which we call conversational localization. To show the feasibility of conversational localization, we develop a proof-of-concept system that guides users to describe their surroundings in a chat and estimates their position based on the information they provide. We devised a modular architecture for our system with four modules. First, we construct an entity database with available image-based floor maps. Second, we enable the dynamic identification and scoring of information provided by users through our utterance processing module. Then, we implement a conversational agent that …

Enhancing 21 U.S.C. §§ 355, 356, And 360 To Encompass Artificial Intelligence-Based Drug Design And Manufacturing Methods, Aj Tsang

Michigan Technology Law Review

Despite newfound attention to how artificial intelligence (AI) may accelerate pharmaceutical development, federal regulators may find that current statutes are ambiguous or silent about their applicability to AI-based drug design and manufacturing methods. This poses a serious problem in the era of Loper Bright and the Major Questions Doctrine. As federal agencies struggle to adjust to courts’ growing demand for Congress to craft clear, explicit, and express delegations of authority, this note develops a statutory framework in which the Food and Drug Administration (FDA) would have more flexibility to regulate the use of AI in advanced drug manufacturing. Guided by …

Judging Our New Judges: Why We Must Remove Artificial Intelligence From Our Courtrooms Now, Kieran Duffy Newcomb

Honors Theses and Capstones

In this paper, I explore some of the ways in which artificial intelligence might enhance the sentencing process through recidivism prediction technology. Notably, this technology can increase the accuracy of risk predictions and the speed with which sentencing decisions are reached. I then show, however, that the recidivism prediction technology is likely to turn into what data scientist Cathy O’Neil calls a Weapon of Math Destruction. The potential harmfulness of this technology is due not to the inherent nature of the technology, but the symbiotic relationship it will have with our already harmful criminal justice system. I argue that the …

The Hazard Prediction Problem, Mary E. Helander, Brendan Smith, Sylvia Charchut, Erika Swiatowy, Calvin Nau, Gregory Cavaretta, Timothy Schuler, Adam Schunk, Héctor Ortiz-Peña

Social Science - All Scholarship

This work formulates the hazard prediction problem while addressing the research question: Can machine learning create a model to automatically recognize patterns that correspond to hazard state conditions during a mission-critical operation? Supervised learning models were trained and tested on data observed from mission simulators, which allowed for safe observation of dynamic system states and undesirable casualty events. The prediction task was formulated as a binary classification problem, producing the probability of being in a hazard state at time t and providing situational awareness of a possible imminent loss. Several modeling architectures were investigated: neural networks, logistic regression, a support …

Chatting With Ai: Deciphering Developer Conversations With Chatgpt, Esteban Parra Rodriguez, Suad Mohamed, Abdullah Parvin

Funded Scholarship

Large Language Models (LLMs) have been widely adopted and are becoming ubiquitous and integral to software development. However, we have little knowledge as to how these tools are being used by software developers beyond anecdotal evidence and word-of-mouth reports. In this work, we present a study toward understanding how developers engage with and utilize LLMs by reporting the results of an empirical study identifying patterns in the conversation that developers have with LLMs. We identified a total of 19 topics describing the purpose of the developers in their conversations with LLMs. Our findings reveal that developers use LLMs to facilitate …