Physical Sciences and Mathematics Commons^™

Communication-Efficient Federated Learning For Leo Constellations Integrated With Haps Using Hybrid Noma-Ofdm, Mohamed Elmahallawy, Tony T. Luo, Khaled Ramadan

Computer Science Faculty Research & Creative Works

Space AI has become increasingly important and sometimes even necessary for government, businesses, and society. An active research topic under this mission is integrating federated learning (FL) with satellite communications (SatCom) so that numerous low Earth orbit (LEO) satellites can collaboratively train a machine learning model. However, the special communication environment of SatCom leads to a very slow FL training process up to days and weeks. This paper proposes NomaFedHAP, a novel FL-SatCom approach tailored to LEO satellites, that (1) utilizes high-altitude platforms (HAPs) as distributed parameter servers (PSs) to enhance satellite visibility, and (2) introduces non-orthogonal multiple access (NOMA) …

Bare-Bones Based Salp Swarm Algorithm For Text Document Clustering, Mohammed Azmi Al-Betar, Ammar Kamal Abasi, Ghazi Al-Naymat, Kamran Arshad, Sharif Naser Makhadmeh

Machine Learning Faculty Publications

Text Document Clustering (TDC) is a challenging optimization problem in unsupervised machine learning and text mining. The Salp Swarm Algorithm (SSA) has been found to be effective in solving complex optimization problems. However, the SSA’s exploitation phase requires improvement to solve the TDC problem effectively. In this paper, we propose a new approach, known as the Bare-Bones Salp Swarm Algorithm (BBSSA), which leverages Gaussian search equations, inverse hyperbolic cosine control strategies, and greedy selection techniques to create new individuals and guide the population towards solving the TDC problem. We evaluated the performance of the BBSSA on six benchmark datasets from …

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 …

Optimal Adaptive Tracking Control Of Partially Uncertain Nonlinear Discrete-Time Systems Using Lifelong Hybrid Learning, Behzad Farzanegan, Rohollah Moghadam, Sarangapani Jagannathan, Pappa Natarajan

Electrical and Computer Engineering Faculty Research & Creative Works

This article addresses a multilayer neural network (MNN)-based optimal adaptive tracking of partially uncertain nonlinear discrete-time (DT) systems in affine form. By employing an actor–critic neural network (NN) to approximate the value function and optimal control policy, the critic NN is updated via a novel hybrid learning scheme, where its weights are adjusted once at a sampling instant and also in a finite iterative manner within the instants to enhance the convergence rate. Moreover, to deal with the persistency of excitation (PE) condition, a replay buffer is incorporated into the critic update law through concurrent learning. To address the vanishing …

Improved Ant Colony Optimization Algorithm For Jamming Resource Allocation, Qingyun Wang, Dezhong Jiao, Shi Shuo, Genyan Peng, Junhua Sun, Yuxin Duan

Journal of System Simulation

Abstract: Ant Colony Optimization (ACO) is a new intelligence optimization algorithm. When applied to jamming resource allocation, the velocity of convergence in optimization process is slow and the probability of obtaining the global optimal solution is low. In order to raise the efficiency of jamming resource allocation and the probability of getting global optimal solution, the attenuation factor is improved to a variable that changes according to the exponential function in optimization process. The attenuation factor is taken as a relatively small value in the initial search phase, and increases monotonically and exponentially as the number of iterations increases. Simulation …

Robot: Robustness-Oriented Testing For Deep Learning Systems, Jingyi Wang, Jialuo Chen, Youcheng Sun, Xingjun Ma, Dongxia Wang, Jun Sun, Peng Cheng

Research Collection School Of Computing and Information Systems

Recently, there has been a significant growth of interest in applying software engineering techniques for the quality assurance of deep learning (DL) systems. One popular direction is deep learning testing, where adversarial examples (a.k.a. bugs) of DL systems are found either by fuzzing or guided search with the help of certain testing metrics. However, recent studies have revealed that the commonly used neuron coverage metrics by existing DL testing approaches are not correlated to model robustness. It is also not an effective measurement on the confidence of the model robustness after testing. In this work, we address this gap by …

Lecture 05: The Convergence Of Big Data And Extreme Computing, David Keyes

Mathematical Sciences Spring Lecture Series

As simulation and analytics enter the exascale era, numerical algorithms, particularly implicit solvers that couple vast numbers of degrees of freedom, must span a widening gap between ambitious applications and austere architectures to support them. We present fifteen universals for researchers in scalable solvers: imperatives from computer architecture that scalable solvers must respect, strategies towards achieving them that are currently well established, and additional strategies currently being developed for an effective and efficient exascale software ecosystem. We consider recent generalizations of what it means to “solve” a computational problem, which suggest that we have often been “oversolving” them at the …

Task Demand Transition Peak Point Effects On Mental Workload Measures Divergence, Enrique Muñoz-De-Escalona, José Juan Cañas, Maria Chiara Leva, Luca Longo

Conference Papers

The capacity to assess and manage mental workload is becoming more and more relevant in the current work environments as it helps to prevent work related accidents and achieve better efficiency and productivity. Mental workload is often measured indirectly by inferring its effects on performance, mental states, and psychophysiological indexes. Since these three main axes should reflect changes in task demands, convergence between measures is expected, however research has found that this convergence is not to be taken for granted as it is not often present. This study aims to explore how the task demand transition peak point may affect …

Modified Sparse Quasi - Newton Algorithm For Electrical Capacitance Tomography System, Chen Yu, Zongji Xia, Yujia Zhou

Journal of System Simulation

Abstract: To solve the 'soft-field' nature and the ill-posed problem in electrical capacitance tomography technology, a modified sparse quasi - newton algorithm for electrical capacitance tomography is presented. The mathematical model of modified sparse quasi - newton is derived. The final iteration formula of modified sparse quasi - newton used to adjust the inverse problem solving of the capacitance tomography image reconstruction is given. An iterative formula for ECT inverse problem solving is used for digital simulation experiment. The simulation experiment results are compared with the results of LBP, Landweber, CG, SD, and so on. The results of the analysis …

The Security Layer, Mark Thomas O'Neill

Theses and Dissertations

Transport Layer Security (TLS) is a vital component to the security ecosystem and the most popular security protocol used on the Internet today. Despite the strengths of the protocol, numerous vulnerabilities result from its improper use in practice. Some of these vulnerabilities arise from weaknesses in authentication, from the rigidity of the trusted authority system to the complexities of client certificates. Others result from the misuse of TLS by developers, who misuse complicated TLS libraries, improperly validate server certificates, employ outdated cipher suites, or deploy other features insecurely. To make matters worse, system administrators and users are powerless to fix …

Optimization Under Fuzzy Constraints: From A Heuristic Algorithm To An Algorithm That Always Converges, Vladik Kreinovich, Juan Carlos Figueroa-Garcia

Departmental Technical Reports (CS)

An efficient iterative heuristic algorithm has been used to implement Bellman-Zadeh solution to the problem of optimization under fuzzy constraints. In this paper, we analyze this algorithm, explain why it works, show that there are cases when this algorithm does not converge, and propose a modification that always converges.

Panorama: Multi-Path Ssl Authentication Using Peer Network Perspectives, William P. Harris

Computer Engineering

SSL currently uses certificates signed by Certificate Authorities (CAs) to authenticate connections. e.g. Google will pay a CA to sign a certificate for them, so that they can prove that they're not someone pretending to be Google. Unfortunately, this system has had multiple problems, and many believe that an alternative needs to be found.

One of the ideas for alternatives is using multiple "network perspectives" to authenticate a server. The idea behind this is that, though playing man-in-the-middle (MITM) with one connection is easy, it should be difficult for an adversary to do so with many connections, especially if they …

Converging And Coexisting Systems Towards Smart Surveillance, Katina Michael, Mg Michael

Professor Katina Michael

Tracking and monitoring people as they operate within their personal networks benefits service providers and their constituents but involves hidden risks and costs.

Automatic identification technologies, CCTV cameras, pervasive and mobile networks, wearable computing, location-based services and social networks have traditionally served distinct purposes. However, we have observed patterns of integration, convergence and coexistence among all these innovations within the information and communication technology industry.1For example, ‘location-based social networking’ can draw on a smart phone's capacity to identify a user uniquely, locate him within 1–2m and share this information across his social network in real time. The resulting ability to …

Momcmc: An Efficient Monte Carlo Method For Multi-Objective Sampling Over Real Parameter Space, Yaohang Li

Computer Science Faculty Publications

In this paper, we present a new population-based Monte Carlo method, so-called MOMCMC (Multi-Objective Markov Chain Monte Carlo). for sampling in the presence of multiple objective functions in real parameter space. The MOMCMC method is designed to address the "multi-objective sampling" problem, which is not only of interest in exploring diversified solutions at the Pareto optimal front in the function space of multiple objective functions, but also those near the front. MOMCMC integrates Differential Evolution (DE) style crossover into Markov Chain Monte Carlo (MCMC) to adaptively propose new solutions from the current population. The significance of dominance is taken into …

Strategic Planning For Digital Convergence In South African Businesses, Manoj Maharaj, Kiru Pillay

Manoj Maharaj

No abstract provided.

Reinforcement Learning Of Competitive And Cooperative Skills In Soccer Agents, Jinsong Leng, Chee Lim

Research outputs 2011

The main aim of this paper is to provide a comprehensive numerical analysis on the efficiency of various reinforcementlearning (RL) techniques in an agent-based soccer game. The SoccerBots is employed as a simulation testbed to analyze the effectiveness of RL techniques under various scenarios. A hybrid agent teaming framework for investigating agent team architecture, learning abilities, and other specific behaviours is presented. Novel RL algorithms to verify the competitiveandcooperativelearning abilities of goal-oriented agents for decision-making are developed. In particular, the tile coding (TC) technique, a function approximation approach, is used to prevent the state space from growing exponentially, hence avoiding …

Planetary-Scale Rfid Services In An Age Of Uberveillance, K. Michael, George Roussos, George Q. Huang, Rajit Gadh, Arunabh Chattopadhyay, S Prabhu, Peter Chu

Professor Katina Michael

Radio-frequency identification has a great number of unfulfilled prospects. Part of the problem until now has been the value proposition behind the technology- it has been marketed as a replacement technique for the barcode when the reality is that it has far greater capability than simply non-line-of-sight identification, towards decision-making in strategic management and reengineered business processes. The vision of the Internet of Things has not eventuated but a world in which every object you can see around you carries the possibility of being connected to the internet is still within the realm of possibility. However incremental innovations may see …

Ocular Vergence Response Over Anaglyphic Stereoscopic Videos, Brian Daugherty

All Theses

The effect of anaglyphic stereographic stimuli on ocular vergence response is examined. An experiment is performed comparing ocular vergence response induced by anaglyphic stereographic display versus standard monoscopic display. Two visualization tools, synchronized three-dimensional scanpath playback and real-time dynamic heatmap generation,
are developed and used to subjectively support the quantitative analysis of ocular disparity. The results of a one-way ANOVA indicate that there is a highly significant effect of anaglyphic stereoscopic display on ocular vergence for a majority of subjects although consistency of vergence response is difficult to predict.

Is It The End Of Barcodes In Supply Chain Management? , Luke Mccathie, Katina Michael

Professor Katina Michael

Barcode is a mature automatic identification (auto-ID) technology that has been used in supply chain management (SCM) for several decades. Such has been the domination of the auto-ID technique that it has pervaded all facets of SCM, from item-level identification to transportation applications. It has enjoyed free reign especially in the retail sector. However, recently radio-frequency identification (RFID) has been considered a rival technology, more superior in terms of its power to store and update information instantaneously, and non-line of sight (nLoS) ability to be read. Yet RFID is more costly and for the present barcode is still the most …

The Hybridization Of Automatic Identification Techniques In Mass Market Applications: Towards A Model Of Coexistence, Katina Michael, M. G. Michael, Holly Tootell, V. Baker

Professor Katina Michael

The number and type of automatic identification technologies in the market have grown since the bar code was introduced in the retail sector in the late 1960s. This paper studies the selection environment of auto-ID and defines, describes and gives examples of three main patterns of innovation: migration, integration, and convergence. The findings indicate that technology adoption is not always about choosing the dominant design but about how to future-proof an auto-ID implementation. Enterprises wishing to adopt auto-ID techniques need to be aware that technology is not static, auto-ID techniques are not stand-alone, and consumers may have wide-ranging requirements for …

Location-Based Services: A Vehicle For It&T Convergence , Katina Michael

Professor Katina Michael

Location-based services (LBS), more than any other mobile commerce application area has served to bring together information technology and telecommunications (IT&T) industries. While much has been written on the potential of LBS, literature on how it is a catalyst for digital convergence is scant. This paper identifies and explores the various levels of converging technologies in mobile commerce by using three LBS case studies. Through literal replication the findings indicate that IT&T technologies are converging at the infrastructure, appliance and application level. It is predicted that mCommerce applications will increasingly rely on industry convergence to achieve their desired outcomes.

Trends In The Selection Of Automatic Identification Technology In Electronic Commerce Applications, Katina Michael

Professor Katina Michael

Since the 1970s, automatic identification (auto-ID) technologies have been evolving to revolutionise the way people live and work. Previous research has not addressed auto-ID technological innovation as a field of study, despite its growing importance on consumer, business and government electronic commerce (EC) applications. This paper is specifically concerned with five auto-ID technologies, bar codes, magnetic-stripe card, smart card, biometrics and radiofrequency identification (RF/ID) tags and transponders. Using multiple embedded case studies and applying the fundamental concepts of the systems of innovation (SI) approach, the overall aim is to understand the selection environment of the auto-ID industry. The results show …

Barriers To Rfid Adoption In The Supply Chain, Nick Huber, Katina Michael, Luke Mccathie

Professor Katina Michael

This paper explores the current barriers to adoption of radio-frequency identification (RFID) for supply chain applications, and documents the perceptions of key players in the Australian RFID market. The paper contains data collected from interviews of both technology providers (e.g. RFID vendors), and prospective business customer (i.e. a large retailer). Data collected is analyzed using qualitative content analysis, and supported with figures and tables. The findings show that the three main barriers to RFID adoption are: the cost of RFID implementation (especially ongoing tag costs), lack of customer awareness and education, and a technology which is only at the beginning …

Limitations And Extensions Of The Wolf-Phc Algorithm, Philip R. Cook

Theses and Dissertations

Policy Hill Climbing (PHC) is a reinforcement learning algorithm that extends Q-learning to learn probabilistic policies for multi-agent games. WoLF-PHC extends PHC with the "win or learn fast" principle. A proof that PHC will diverge in self-play when playing Shapley's game is given, and WoLF-PHC is shown empirically to diverge as well. Various WoLF-PHC based modifications were created, evaluated, and compared in an attempt to obtain convergence to the single shot Nash equilibrium when playing Shapley's game in self-play without using more information than WoLF-PHC uses. Partial Commitment WoLF-PHC (PCWoLF-PHC), which performs best on Shapley's game, is tested on other …

Fuzzy Membership Function Initial Values: Comparing Initialization Methods That Expedite Convergence, Stephanie Scheibe Lee

Theses and Dissertations

Fuzzy attributes are used to quantify imprecise data that model real world objects. To effectively use fuzzy attributes, a fuzzy membership function must be defined to provide the boundaries for the fuzzy data. The initialization of these membership function values should allow the data to converge to a stable membership value in the shortest time possible. The paper compares three initialization methods, Random, Midpoint and Random Proportional, to determine which method optimizes convergence. The comparison experiments suggest the use of the Random Proportional method.

Simple Genetic Algorithms With Linear Fitness, Michael D. Vose, Alden H. Wright

Computer Science Faculty Publications

A general form of stochastic search is described (random heuristic search), and some of its general properties are proved. This provides a framework in which the simple genetic algorithm (SGA) is a special case. The framework is used to illuminate relationships between seemingly different probabilistic perspectives of SGA behavior. Next, the SGA is formalized as an instance of random heuristic search. The formalization then used to show expected population fitness is a Lyapunov function in the infinite population model when mutation is zero and fitness is linear. In particular, the infinite population algorithm must converge, and average population …

Parallel Error Tolerance Scheme Based On The Hill Climbing Nature Of Simulated Annealing, Bruce M. Mcmillin, Chul-Eui Hong

Computer Science Faculty Research & Creative Works

In parallelizing simulated annealing in a multicomputer, maintaining the global state S involves explicit message traffic and is a critical performance bottleneck. One way to mitigate this bottleneck is to amortize the overhead of these state updates over as many parallel state changes as possible. Using this technique introduces errors in the calculated cost C(S) of a particular state S used by the annealing process. Analytically derived bounds are placed on this error in order to assure convergence to the correct result. The resulting parallel simulated annealing algorithm dynamically changes the frequency of global updates as a function of the …

Parallel Implementation Of A Recursive Least Squares Neural Network Training Method On The Intel Ipsc/2, James Edward Steck, Bruce M. Mcmillin, K. Krishnamurthy, M. Reza Ashouri, Gary G. Leininger

Computer Science Faculty Research & Creative Works

An algorithm based on the Marquardt-Levenberg least-square optimization method has been shown by S. Kollias and D. Anastassiou (IEEE Trans. on Circuits Syst. vol.36, no.8, p.1092-101, Aug. 1989) to be a much more efficient training method than gradient descent, when applied to some small feedforward neural networks. Yet, for many applications, the increase in computational complexity of the method outweighs any gain in learning rate obtained over current training methods. However, the least-squares method can be more efficiently implemented on parallel architectures than standard methods. This is demonstrated by comparing computation times and learning rates for the least-squares method implemented …

Algorithms For Pipe Network Analysis And Their Reliability, Don J. Wood

KWRRI Research Reports

Algorithms for analyzing steady state flow conditions in pipe networks are developed for general applications. The algorithms are based on both loop equations expressed in terms of unknown flowrates and node equations expressed in terms of unknown grades. Five methods, which represent those in significant use today, are presented. An example pipe network is analyzed to illustrate the application of the various algorithms. The various assumptions required for the different methods are presented and the methods are compared within a common framework.

The reliabilities of these commonly employed algorithms for pipe network analysis are investigated by analyzing a large number …