Physical Sciences and Mathematics Commons^™

System Identification Through Lipschitz Regularized Deep Neural Networks, Elisa Negrini, Giovanna Citti, Luca Capogna

Mathematics Sciences: Faculty Publications

In this paper we use neural networks to learn governing equations from data. Specifically we reconstruct the right-hand side of a system of ODEs x˙(t)=f(t,x(t)) directly from observed uniformly time-sampled data using a neural network. In contrast with other neural network-based approaches to this problem, we add a Lipschitz regularization term to our loss function. In the synthetic examples we observed empirically that this regularization results in a smoother approximating function and better generalization properties when compared with non-regularized models, both on trajectory and non-trajectory data, especially in presence of noise. In contrast with sparse regression approaches, since neural networks …

Binary Classifiers For Noisy Datasets: A Comparative Study Of Existing Quantum Machine Learning Frameworks And Some New Approaches, Nikolaos Schetakis, Davit Aghamalyan, Paul Robert Griffin, Michael Boguslavsky

Research Collection School Of Computing and Information Systems

This technology offer is a quantum machine learning algorithm applied to binary classification models for noisy datasets which are prevalent in financial and other datasets. By combining hybrid-neural networks, quantum parametric circuits, and data re-uploading we have improved the classification of non-convex 2-dimensional figures by understanding learning stability as noise increases in the dataset. The metric we use for assessing the performance of our quantum classifiers is the area under the receiver operator curve (ROC AUC). We are interested to collaborate with partners with use cases for binary classification of noisy data. Also, as quantum technology is still insufficient for …

Exploratory Data Mining Techniques (Decision Tree Models) For Examining The Impact Of Internet-Based Cognitive Behavioral Therapy For Tinnitus: Machine Learning Approach, Hansapani Rodrigo, Eldré W. Beukes, Gerhard Andersson, Vinaya Manchaiah

School of Mathematical and Statistical Sciences Faculty Publications and Presentations

Background: There is huge variability in the way that individuals with tinnitus respond to interventions. These experiential variations, together with a range of associated etiologies, contribute to tinnitus being a highly heterogeneous condition. Despite this heterogeneity, a “one size fits all” approach is taken when making management recommendations. Although there are various management approaches, not all are equally effective. Psychological approaches such as cognitive behavioral therapy have the most evidence base. Managing tinnitus is challenging due to the significant variations in tinnitus experiences and treatment successes. Tailored interventions based on individual tinnitus profiles may improve outcomes. Predictive models of treatment …

Building Legal Datasets, Jerrold Soh

Research Collection Yong Pung How School Of Law

Data-centric AI calls for better, not just bigger, datasets. As data protection laws with extra-territorial reach proliferate worldwide, ensuring datasets are legal is an increasingly crucial yet overlooked component of “better”. To help dataset builders become more willing and able to navigate this complex legal space, this paper reviews key legal obligations surrounding ML datasets, examines the practical impact of data laws on ML pipelines, and offers a framework for building legal datasets.

Multi-Object Localization In Robotic Hand, Tsing Tsow

USF Tampa Graduate Theses and Dissertations

We have developed a machine learning approach to localized objects inside a robotic hand using only images from 2D cameras. Specifically, we used deep learning method (You Only Look Once, YOLO) and Iterative closest Point (ICP) to estimate the 3D coordinates of the objects in a robotic hand. This method will also output the number of objects inside the robotic hand in addition to the coordinates of the objects. We have demonstrated the performance with simulation and obtained typical accuracy within a few pixels (couple mm) and counting accuracy of about 76%. We have also applied it to real images, …

Towards Accurate Run-Time Hardware-Assisted Stealthy Malware Detection: A Lightweight, Yet Effective Time Series Cnn-Based Approach, Hossein Sayadi, Yifeng Gao, Hosein Mohammadi Makrani, Jessica Lin, Paulo Cesar Costa, Setareh Rafatirad, Houman Homayoun

Computer Science Faculty Publications and Presentations

According to recent security analysis reports, malicious software (a.k.a. malware) is rising at an alarming rate in numbers, complexity, and harmful purposes to compromise the security of modern computer systems. Recently, malware detection based on low-level hardware features (e.g., Hardware Performance Counters (HPCs) information) has emerged as an effective alternative solution to address the complexity and performance overheads of traditional software-based detection methods. Hardware-assisted Malware Detection (HMD) techniques depend on standard Machine Learning (ML) classifiers to detect signatures of malicious applications by monitoring built-in HPC registers during execution at run-time. Prior HMD methods though effective have limited their study on …

Diffusional Fractionation Of Helium Isotopes In Silicate Melts, Haiyang Luo, Bijaya Karki, Dipta B. Ghosh, Huiming Bao

Faculty Publications

Estimating Helium (He) concentration and isotope composition of the mantle requires quantifying He loss during magma degassing. The knowledge of diffusional He isotope fractionation in silicate melts may be essential to constrain the He loss. Isotopic mass dependence of He diffusion can be empirically expressed as D3He/D4He = (4/3)^β, where D is the diffusivity of a He isotope. However, no studies have reported any β values for He in silicate melts due to technical challenges in both experiments and computations. Here, molecular dynamics simulations based on deep neural network potentials trained by ab initio data …

Statistical Potentials For Rna-Protein Interactions Optimized By Cma-Es, Takayuki Kimura, Nobuaki Yasuo, Masakazu Sekijima, Brooke Lustig

Faculty Research, Scholarly, and Creative Activity

Characterizing RNA-protein interactions remains an important endeavor, complicated by the difficulty in obtaining the relevant structures. Evaluating model structures via statistical potentials is in principle straight-forward and effective. However, given the relatively small size of the existing learning set of RNA-protein complexes optimization of such potentials continues to be problematic. Notably, interaction-based statistical potentials have problems in addressing large RNA-protein complexes. In this study, we adopted a novel strategy with covariance matrix adaptation (CMA-ES) to calculate statistical potentials, successfully identifying native docking poses.

Deep Learning Applications In Medical Bioinformatics, Ziad Omar

Electronic Theses and Dissertations

After a patient’s breast cancer diagnosis, identifying breast cancer lymph node metastases is one of the most important and critical factor that is directly related to the patient’s survival. The traditional way to examine the existence of cancer cells in the breast lymph nodes is through a lymph node procedure, biopsy. The procedure process is time-consuming for the patient and the provider, costly, and lacks accuracy as not every lymph node is examined. The intent of this study is to develop an artificial neural network (ANNs) that would map genetic biomarkers to breast lymph node classes using ANNs. The neural …

Measuring Data Collection Diligence For Community Healthcare, Galawala Ramesha Samurdhi Karunasena, M. S. Ambiya, Arunesh Sinha, R. Nagar, S. Dalal, Abdullah. H., D. Thakkar, D. Narayanan, M. Tambe

Research Collection School Of Computing and Information Systems

Data analytics has tremendous potential to provide targeted benefit in low-resource communities, however the availability of highquality public health data is a significant challenge in developing countries primarily due to non-diligent data collection by community health workers (CHWs). Our use of the word non-diligence here is to emphasize that poor data collection is often not a deliberate action by CHW but arises due to a myriad of factors, sometime beyond the control of the CHW. In this work, we define and test a data collection diligence score. This challenging unlabeled data problem is handled by building upon domain expert’s guidance …

Comparative Study Of Reinforcement Learning Methods In Path Planning, Daniel Obawole

Electronic Theses and Dissertations

In order to perform a large variety of tasks and achieve human-level performance in complex real-world environments, an intelligent agent must be able to learn from its dynamically changing environment. Generally speaking, agents have limitations in obtaining an accurate description of the environment from what they perceive because they may not have all the information about the environment. The present research is focused on reinforcement learning algorithms that represent a defined category in the field of machine learning because of their unique approach based on a trial-error basis. Reinforcement learning is used to solve control problems based on received rewards. …

Deep Fakes: The Algorithms That Create And Detect Them And The National Security Risks They Pose, Nick Dunard

James Madison Undergraduate Research Journal (JMURJ)

The dissemination of deep fakes for nefarious purposes poses significant national security risks to the United States, requiring an urgent development of technologies to detect their use and strategies to mitigate their effects. Deep fakes are images and videos created by or with the assistance of AI algorithms in which a person’s likeness, actions, or words have been replaced by someone else’s to deceive an audience. Often created with the help of generative adversarial networks, deep fakes can be used to blackmail, harass, exploit, and intimidate individuals and businesses; in large-scale disinformation campaigns, they can incite political tensions around the …

Artificial Intelligence In Cyber Security, University Of Maine Artificial Intelligence Initiative

General University of Maine Publications

UMaine AI draws top talent and leverages a distinctive set of capabilities from the University of Maine and other collaborating institutions from across Maine and beyond, while it also recruits world-class talent from across the nation and the world. It is centered at the University of Maine, leveraging the university’s strengths across disciplines, including computing and information sciences, engineering, health and life sciences, business, education, social sciences, and more.

Cognition-Enhanced Machine Learning For Better Predictions With Limited Data, Florian Sense, Ryan Wood, Michael G. Collins, Joshua Fiechter, Aihua W. Wood, Michael Krusmark, Tiffany Jastrzembski, Christopher W. Myers

Faculty Publications

The fields of machine learning (ML) and cognitive science have developed complementary approaches to computationally modeling human behavior. ML's primary concern is maximizing prediction accuracy; cognitive science's primary concern is explaining the underlying mechanisms. Cross-talk between these disciplines is limited, likely because the tasks and goals usually differ. The domain of e-learning and knowledge acquisition constitutes a fruitful intersection for the two fields’ methodologies to be integrated because accurately tracking learning and forgetting over time and predicting future performance based on learning histories are central to developing effective, personalized learning tools. Here, we show how a state-of-the-art ML model can …

Critical Behavior In Evolutionary And Population Dynamics, Stephen Ordway

Dissertations

This study is an exploration of phase transition behavior in evolutionary and population dynamics, and techniques for predicting population changes, across the disciplines of physics, biology, and computer science. Under the looming threat of climate change, it is imperative to understand the dynamics of populations under environmental stress and to identify early warning signals of population decline. These issues are explored here in (1) a computational model of evolutionary dynamics, (2) an experimental system of decaying populations under environmental stress, and (3) a machine learning approach to predict population changes based on environmental factors. Through the lens of critical phase …

Exploiting Group Structures To Infer Social Interactions From Videos, Maksim Bolonkin

Dartmouth College Ph.D Dissertations

In this thesis, we consider the task of inferring the social interactions between humans by analyzing multi-modal data. Specifically, we attempt to solve some of the problems in interaction analysis, such as long-term deception detection, political deception detection, and impression prediction. In this work, we emphasize the importance of using knowledge about the group structure of the analyzed interactions. Previous works on the matter mostly neglected this aspect and analyzed a single subject at a time. Using the new Resistance dataset, collected by our collaborators, we approach the problem of long-term deception detection by designing a class of histogram-based features …

Improving Animal Monitoring Using Small Unmanned Aircraft Systems (Suas) And Deep Learning Networks, Meilun Zhou, Jared A. Elmore, Sathishkumar Samiappan, Kristine O. Evans, Morgan Pfeiffer, Bradley F. Blackwell, Raymond B. Iglay

United States Department of Agriculture Wildlife Services: Staff Publications

In recent years, small unmanned aircraft systems (sUAS) have been used widely to monitor animals because of their customizability, ease of operating, ability to access difficult to navigate places, and potential to minimize disturbance to animals. Automatic identification and classification of animals through images acquired using a sUAS may solve critical problems such as monitoring large areas with high vehicle traffic for animals to prevent collisions, such as animal-aircraft collisions on airports. In this research we demonstrate automated identification of four animal species using deep learning animal classification models trained on sUAS collected images. We used a sUAS mounted with …

Characterizing Convolutional Neural Network Early-Learning And Accelerating Non-Adaptive, First-Order Methods With Localized Lagrangian Restricted Memory Level Bundling, Benjamin O. Morris

Theses and Dissertations

This dissertation studies the underlying optimization problem encountered during the early-learning stages of convolutional neural networks and introduces a training algorithm competitive with existing state-of-the-art methods. First, a Design of Experiments method is introduced to systematically measure empirical second-order Lipschitz upper bound and region size estimates for local regions of convolutional neural network loss surfaces experienced during the early-learning stages. This method demonstrates that architecture choices can significantly impact the local loss surfaces traversed during training. Next, a Design of Experiments method is used to study the effects convolutional neural network architecture hyperparameters have on different optimization routines' abilities to …

Orthogonal Inductive Matrix Completion, Antoine Ledent, Rrodrigo Alves, Marius Kloft

Research Collection School Of Computing and Information Systems

We propose orthogonal inductive matrix completion (OMIC), an interpretable approach to matrix completion based on a sum of multiple orthonormal side information terms, together with nuclear-norm regularization. The approach allows us to inject prior knowledge about the singular vectors of the ground-truth matrix. We optimize the approach by a provably converging algorithm, which optimizes all components of the model simultaneously. We study the generalization capabilities of our method in both the distribution-free setting and in the case where the sampling distribution admits uniform marginals, yielding learning guarantees that improve with the quality of the injected knowledge in both cases. As …

Novel Statistical Modeling Methods For Traffic Video Analysis, Hang Shi

Dissertations

Video analysis is an active and rapidly expanding research area in computer vision and artificial intelligence due to its broad applications in modern society. Many methods have been proposed to analyze the videos, but many challenging factors remain untackled. In this dissertation, four statistical modeling methods are proposed to address some challenging traffic video analysis problems under adverse illumination and weather conditions.

First, a new foreground detection method is presented to detect the foreground objects in videos. A novel Global Foreground Modeling (GFM) method, which estimates a global probability density function for the foreground and applies the Bayes decision rule …

Gradient Free Sign Activation Zero One Loss Neural Networks For Adversarially Robust Classification, Yunzhe Xue

Dissertations

The zero-one loss function is less sensitive to outliers than convex surrogate losses such as hinge and cross-entropy. However, as a non-convex function, it has a large number of local minima, andits undifferentiable attribute makes it impossible to use backpropagation, a method widely used in training current state-of-the-art neural networks. When zero-one loss is applied to deep neural networks, the entire training process becomes challenging. On the other hand, a massive non-unique solution probably also brings different decision boundaries when optimizing zero-one loss, making it possible to fight against transferable adversarial examples, which is a common weakness in deep learning …

Towards Adversarial Robustness With 01 Lossmodels, And Novel Convolutional Neural Netsystems For Ultrasound Images, Meiyan Xie

Dissertations

This dissertation investigates adversarial robustness with 01 loss models and a novel convolutional neural net systems for vascular ultrasound images.

In the first part, the dissertation presents stochastic coordinate descent for 01 loss and its sensitivity to adversarial attacks. The study here suggests that 01 loss may be more resilient to adversarial attacks than the hinge loss and further work is required.

In the second part, this dissertation proposes sign activation network with a novel gradient-free stochastic coordinate descent algorithm and its ensembling model. The study here finds that the ensembling model gives a high minimum distortion (as measured by …

Data-Driven Learning For Robot Physical Intelligence, Leidi Zhao

Dissertations

The physical intelligence, which emphasizes physical capabilities such as dexterous manipulation and dynamic mobility, is essential for robots to physically coexist with humans. Much research on robot physical intelligence has achieved success on hyper robot motor capabilities, but mostly through heavily case-specific engineering. Meanwhile, in terms of robot acquiring skills in a ubiquitous manner, robot learning from human demonstration (LfD) has achieved great progress, but still has limitations handling dynamic skills and compound actions. In this dissertation, a composite learning scheme which goes beyond LfD and integrates robot learning from human definition, demonstration, and evaluation is proposed. This method tackles …

Evaluating Machine Learning Model Stability For Software Bug Prediction, Joud El-Shawa

Undergraduate Student Research Internships Conference

Large software systems are implemented using many different programming languages and scripts, and consequently the dependencies between their components are very complex. It is therefore difficult to extract and understand these dependencies by solely analyzing the source code, so that failure risks can be detected accurately. On the other hand, it is a common practice for software engineers to keep track of process related metrics such as the number of times a component was maintained, with which other components it has been co-committed, whether the maintenance activity was a bug-fixing activity, and how many lines of source code have been …

Turn Of Phrase: Contrastive Pre-Training For Discourse-Aware Conversation Models, Roland Laboulaye

Theses and Dissertations

Understanding long conversations requires recognizing a discourse flow unique to conversation. Recent advances in unsupervised representation learning of text have been attained primarily through language modeling, which models discourse only implicitly and within a small window. These representations are in turn evaluated chiefly on sentence pair or paragraph-question pair benchmarks, which measure only local discourse coherence. In order to improve performance on discourse-reliant, long conversation tasks, we propose Turn-of-Phrase pre-training, an objective designed to encode long conversation discourse flow. We leverage tree-structured Reddit conversations in English to, relative to a chosen conversation path through the tree, select paths of varying …

Machine Learning For Analog/Mixed-Signal Integrated Circuit Design Automation, Weidong Cao

McKelvey School of Engineering Theses & Dissertations

Analog/mixed-signal (AMS) integrated circuits (ICs) play an essential role in electronic systems by processing analog signals and performing data conversion to bridge the analog physical world and our digital information world.Their ubiquitousness powers diverse applications ranging from smart devices and autonomous cars to crucial infrastructures. Despite such critical importance, conventional design strategies of AMS circuits still follow an expensive and time-consuming manual process and are unable to meet the exponentially-growing productivity demands from industry and satisfy the rapidly-changing design specifications from many emerging applications. Design automation of AMS IC is thus the key to tackling these challenges and has been …

Exploratory Search With Archetype-Based Language Models, Brent D. Davis

Electronic Thesis and Dissertation Repository

This dissertation explores how machine learning, natural language processing and information retrieval may assist the exploratory search task. Exploratory search is a search where the ideal outcome of the search is unknown, and thus the ideal language to use in a retrieval query to match it is unavailable. Three algorithms represent the contribution of this work. Archetype-based Modeling and Search provides a way to use previously identified archetypal documents relevant to an archetype to form a notion of similarity and find related documents that match the defined archetype. This is beneficial for exploratory search as it can generalize beyond standard …

Comparative Study Of Machine Learning Models On Solar Flare Prediction Problem, Nikhil Sai Kurivella

All Graduate Theses and Dissertations, Spring 1920 to Summer 2023

Solar flare events are explosions of energy and radiation from the Sun’s surface. These events occur due to the tangling and twisting of magnetic fields associated with sunspots. When Coronal Mass ejections accompany solar flares, solar storms could travel towards earth at very high speeds, disrupting all earthly technologies and posing radiation hazards to astronauts. For this reason, the prediction of solar flares has become a crucial aspect of forecasting space weather. Our thesis utilized the time-series data consisting of active solar region magnetic field parameters acquired from SDO that span more than eight years. The classification models take AR …

Applying Deep Learning To The Ice Cream Vendor Problem: An Extension Of The Newsvendor Problem, Gaffar Solihu

Electronic Theses and Dissertations

The Newsvendor problem is a classical supply chain problem used to develop strategies for inventory optimization. The goal of the newsvendor problem is to predict the optimal order quantity of a product to meet an uncertain demand in the future, given that the demand distribution itself is known. The Ice Cream Vendor Problem extends the classical newsvendor problem to an uncertain demand with unknown distribution, albeit a distribution that is known to depend on exogenous features. The goal is thus to estimate the order quantity that minimizes the total cost when demand does not follow any known statistical distribution. The …

Two Dimensional Nanoparticles/Nanozymes For Biosensing Augmented By Machine Learning, Nidhi Chintan Nandu

Legacy Theses & Dissertations (2009 - 2024)

Biosensing is an ever-evolving field with many resources devoted towards new approaches in sensing and refining the existing methods. For a long time, such approaches involved use of state-of-the-art instruments and were often proved to be time consuming and expensive. Nanoparticles with their unique properties, inexpensive synthesis and ease of manipulation have found purpose in myriad fields including biosensing. Different nanoparticles based on their composition, morphology, dimensionality, and surface modifications have already gained foot hold in the biosensor ranks. The unique opto-electric properties due to the quantum effect in the nanoparticles have also made them one of the leading alternatives …