Physical Sciences and Mathematics Commons^™

Towards An Automated Weight Lifting Coach: Introducing Lift, Michael Andrew Lady

Master's Theses

The fitness device market is young and rapidly growing. More people than ever before take count of how many steps they walk, how many calories they burn, their heart rate over time, and even their quality of sleep. New, and as of yet, unreleased fitness devices have promised the next evolution of functionality with exercise technique analysis. These next generation of fitness devices have wrist and armband style form factors, which may not be optimal for barbell exercises such as back squat, bench press, and overhead press where a sensor on one arm may not provide the most relevant data …

Revisiting Risk-Sensitive Mdps: New Algorithms And Results, Ping Hou, William Yeoh, Pradeep Reddy Varakantham

Research Collection School Of Computing and Information Systems

While Markov Decision Processes (MDPs) have been shown to be effective models for planning under uncertainty, theobjective to minimize the expected cumulative cost is inappropriate for high-stake planning problems. As such, Yu, Lin, and Yan (1998) introduced the Risk-Sensitive MDP (RSMDP) model, where the objective is to find a policy that maximizes the probability that the cumulative cost is within some user-defined cost threshold. In this paper, we revisit this problem and introduce new algorithms that are based on classical techniques, such as depth-first search and dynamic programming, and a recently introduced technique called Topological Value Iteration (TVI). We demonstrate …

Lexical Based Semantic Orientation Of Online Customer Reviews And Blogs-J-Am Sci 10(8) 143_147--07-June-2014.Pdf, Dr. Muhammad Zubair Asghar

Dr. Muhammad Zubair Asghar

Rapid increase in internet users along with growing power of online review sites and social media hasgiven birth to sentiment analysis or opinion mining, which aims at determining what other people think andcomment. Sentiments or Opinions contain public generated content about products, services, policies and politics.People are usually interested to seek positive and negative opinions containing likes and dislikes, shared by users forfeatures of particular product or service. This paper proposed sentence-level lexical based domain independentsentiment classification method for different types of data such as reviews and blogs. The proposed method is basedon general lexicons i.e. WordNet, SentiWordNet and user …

Procedural-Reasoning Architecture For Applied Behavior Analysis-Based Instructions, Edmon Begoli

Doctoral Dissertations

Autism Spectrum Disorder (ASD) is a complex developmental disability affecting as many as 1 in every 88 children. While there is no known cure for ASD, there are known behavioral and developmental interventions, based on demonstrated efficacy, that have become the predominant treatments for improving social, adaptive, and behavioral functions in children.

Applied Behavioral Analysis (ABA)-based early childhood interventions are evidence based, efficacious therapies for autism that are widely recognized as effective approaches to remediation of the symptoms of ASD. They are, however, labor intensive and consequently often inaccessible at the recommended levels.

Recent advancements in socially assistive robotics and …

A Continuous Learning Strategy For Self-Organizing Maps Based On Convergence Windows, Gregory T. Breard

Senior Honors Projects

A self-organizing map (SOM) is a type of artificial neural network that has applications in a variety of fields and disciplines. The SOM algorithm uses unsupervised learning to produce a low-dimensional representation of high- dimensional data. This is done by 'fitting' a grid of nodes to a data set over a fixed number of iterations. With each iteration, the nodes of the map are adjusted so that they appear more like the data points. The low-dimensionality of the resulting map means that it can be presented graphically and be more intuitively interpreted by humans. However, it is still essential to …

Decentralized Multi-Agent Reinforcement Learning In Average-Reward Dynamic Dcops, Duc Thien Nguyen, William Yeoh, Hoong Chuin Lau, Shlomo Zilberstein

Research Collection School Of Computing and Information Systems

Researchers have introduced the Dynamic Distributed Constraint Optimization Problem (Dynamic DCOP) formulation to model dynamically changing multi-agent coordination problems, where a dynamic DCOP is a sequence of (static canonical) DCOPs, each partially different from the DCOP preceding it. Existing work typically assumes that the problem in each time step is decoupled from the problems in other time steps, which might not hold in some applications. Therefore, in this paper, we make the following contributions: (i) We introduce a new model, called Markovian Dynamic DCOPs (MD-DCOPs), where the DCOP in the next time step is a function of the value assignments …

Mechanisms For Arranging Ride Sharing And Fare Splitting For Last-Mile Travel Demands, Shih-Fen Cheng, Duc Thien Nguyen, Hoong Chuin Lau

Research Collection School Of Computing and Information Systems

A great challenge of city planners is to provide efficient and effective connection service to travelers using public transportation system. This is commonly known as the last-mile problem and is critical in promoting the utilization of public transportation system. In this paper, we address the last-mile problem by considering a dynamic and demand-responsive mechanism for arranging ride sharing on a non-dedicated commercial fleet (such as taxis or passenger vans). Our approach has the benefits of being dynamic, flexible, and with low setup cost. A critical issue in such ride-sharing service is how riders should be grouped and serviced, and how …

On Coordinating Pervasive Persuasive Agents, Budhitama Subagdja, Ah-Hwee Tan

Research Collection School Of Computing and Information Systems

There is a growing interest in applying multiagent systems for smart-home environment supporting self-caring elderly. In this paper we investigate situations and conditions for coordination for such kind of system. We specify a high level architecture of it based on the notions of beliefs, desires, and intentions for both individual and group behavior of the agents including the human occupant's. The framework enables flexible coordinations among loosely-coupled heterogeneous agents that converse with the user. This work is conducted towards producing a coordination framework for agents and people in such a kind of smart-home environment as mentioned.

Declarative-Procedural Memory Interaction In Learning Agents, Wenwen Wang, Ah-Hwee Tan, Loo-Nin Teow, Tan Yuan-Sin

Research Collection School Of Computing and Information Systems

It has been well recognized that human makes use of both declarative memory and procedural memory for decision making and problem solving. In this paper, we propose a computational model with the overall architecture and individual processes for realizing the interaction between the declarative and procedural memory based on self-organizing neural networks. We formalize two major types of memory interactions and show how each of them can be embedded into autonomous reinforcement learning agents. Our experiments based on the Toad and Frog puzzle and a strategic game known as Starcraft Broodwar have shown that the cooperative interaction between declarative knowledge …

On Understanding Diffusion Dynamics Of Patrons At A Theme Park, Jiali Du, Akshat Kumar, Pradeep Reddy Varakantham

Research Collection School Of Computing and Information Systems

In this work, we focus on the novel application of learning the diffusion dynamics of visitors among attractions at a large theme park using only aggregate information about waiting times at attractions. Main contributions include formulating optimisation models to compute diffusion dynamics. We also developed algorithm capable of dealing with noise in the data to populate parameters in the optimization model. We validated our approach using cross validation on a real theme park data set. Our approach provides an accuracy of about 80$% for popular attractions, providing solid empirical support for our diffusion models.

A Quantitative Analysis Of Decision Process In Social Groups Using Human Trajectories, Truc Viet Le, Siyuan Liu, Hoong Chuin Lau, Ramayya Krishnan

Research Collection School Of Computing and Information Systems

A group's collective action is an outcome of the group's decision-making process, which may be reached by either averaging of the individual preferences or following the choices of certain members in the group. Our problem here is to decide which decision process the group has adopted given the data of the collective actions. We propose a generic statistical framework to infer the group's decision process from the spatio-temporal data of group trajectories, where each "trajectory" is a sequence of group actions. This is achieved by systematically comparing each agent type's influence on the group actions based on an array of …

Deep Learning Via Stacked Sparse Autoencoders For Automated Voxel-Wise Brain Parcellation Based On Functional Connectivity, Céline Gravelines

Electronic Thesis and Dissertation Repository

Functional brain parcellation – the delineation of brain regions based on functional connectivity – is an active research area lacking an ideal subject-specific solution independent of anatomical composition, manual feature engineering, or heavily labelled examples. Deep learning is a cutting-edge area of machine learning on the forefront of current artificial intelligence developments. Specifically, autoencoders are artificial neural networks which can be stacked to form hierarchical sparse deep models from which high-level features are compressed, organized, and extracted, without labelled training data, allowing for unsupervised learning. This thesis presents a novel application of stacked sparse autoencoders to the problem of parcellating …

Investigating Genotype-Phenotype Relationship Extraction From Biomedical Text, Maryam Khordad

Electronic Thesis and Dissertation Repository

During the last decade biomedicine has developed at a tremendous pace. Every day a lot of biomedical papers are published and a large amount of new information is produced. To help enable automated and human interaction in the multitude of applications of this biomedical data, the need for Natural Language Processing systems to process the vast amount of new information is increasing. Our main purpose in this research project is to extract the relationships between genotypes and phenotypes mentioned in the biomedical publications. Such a system provides important and up-to-date data for database construction and updating, and even text summarization. …

Learning Two-Input Linear And Nonlinear Analog Functions With A Simple Chemical System, Peter Banda, Christof Teuscher

Computer Science Faculty Publications and Presentations

The current biochemical information processing systems behave in a predetermined manner because all features are defined during the design phase. To make such unconventional computing systems reusable and programmable for biomedical applications, adaptation, learning, and self-modification baaed on external stimuli would be highly desirable. However, so far, it haa been too challenging to implement these in real or simulated chemistries. In this paper we extend the chemical perceptron, a model previously proposed by the authors, to function as an analog instead of a binary system. The new analog asymmetric signal perceptron learns through feedback and supports MichaelisMenten kinetics. The results …

Of Mills And Machines - Computing Thought Experiments On Consciousness, Aïda Raoult

Aïda Raoult

In this computing-oriented analysis, 5 of the most influential thought experiments on consciousness inspired by the development of A.I. in the 70s-80s are (1) presented as more refined versions of Leibniz’s mill (LM), (2) then reformulated in terms of LM which reveals a divergence in their approach of the mind-body problem; (3) combining this result with computational complexity theory shows the ontological question is less difficult to answer than the causal one. In the end, (4) these considerations participate in the debate over machine consciousness.

Constraint Satisfaction Problem: A Generic Scheduler, Ben Carpenter, Brent Weichel, Jeremy Straub, Eunjin Kim

Jeremy Straub

The task was to create a scheduler that would create a schedule that gets as many of the tasks done as possible while maximizing the total value of the tasks performed. Each task was assigned a value, a priority, and a duration. Each task also had certain times that they could be run, so they couldn’t just be run at any point where they fit. We decided that in order to get a more accurate ordering for the process, we would take the value divided by the duration that way we were less likely to skip over processes that ran …

Dynamic Task Scheduling Problem: Greedy Knapsack Solution, Christian Sandtveit, Darrin Winger, Jeremy Straub, Eunjin Kim

Jeremy Straub

The problem that we worked with was a dynamic scheduling problem. For this problem, we are given a set of tasks to be scheduled in an allotted time slot, so that the total value of the tasks done is maximized. Each task has a duration, value. Each task also has one or more periods in which they can be scheduled. Some tasks can have conflicting time slots that can prevent other tasks from being scheduled. As tasks are assigned time slots it is possible to prevent other tasks from being as-signed a time slot. Looking for ways to minimize the …

Medical Rate Setting: Multi-Curve Approximation And Projection, Darrin Winger, Christian Sandtveit, Jeremy Straub, Eunjin Kim

Jeremy Straub

In order to maximize profit, our approach was to maximize the difference between total revenue and total cost, where total revenue would be larger than total cost. In the problem we are given a series of points, which relates price, cost, profit and quantity. We can calculate the total revenue by multi-plying the price with quantity, and the total cost by multiplying the cost with the quantity. Total profit is calculated by multiplying profit and quantity. We are given 4 initial points, and based on those 4 points we will calculate the point where the profit is currently maximized. Based …

Task Scheduling Problem: Using The Most Constrained Variable Algorithm To Maximize, Jaeden Lovin, Calvin Bina, Jeremy Straub, Eunjin Kim

Jeremy Straub

For this constraint satisfaction problem we needed to schedule a series of tasks to run in a certain order. Each task has a set duration that it must run for and a domain of times during which it can run during. Each task had a value and the goal of the problem was to pick times for the tasks to run in or-der to maximize the total value. We thought of multiple ways to potentially approach this problem, and decided to use some form of the least constraining variable. We would choose the task with the least constraints on other …

Medical Rate Setting Problem: Using The Hill-Climbing Search To Maximize Health Care Provider Profit, Calvin Bina, Jaeden Lovin, Jeremy Straub, Eunjin Kim

Jeremy Straub

Our program for calculating the optimal price for a service is relatively simple, but it gets great results. We make use of quadratic regres-sion. Quadratic regression has a very similar concept to linear regression. Given a set of data points, we find the equation that is the best fit to represent those data points. With linear re-gression, our resulting equation is linear. How-ever, with quadratic regression, our end result is a quadratic equation. We have two quadratic equations to come up with. One is our cost function and the other is our units sold func-tion. Both of these equations are …

Update On The Operating Software For Openorbiter, Dayln Limesand, Christoffer Korvald, Jeremy Straub, Ronald Marsh

Jeremy Straub

The operating software team of the OpenOrbiter project has been tasked with developing software for general spacecraft maintenance, performing mission tasks and the monitoring of system critical aspects of the spacecraft. To do so, the team is developing an autonomous system that will be able to continuously check sensors for data, and schedule tasks that pertain to the current mission and general maintenance of the onboard systems. Development in support of these objectives is ongoing with work focusing on the completion of the development of a stable system. This poster presents an overview of current work on the project and …

Moving Object Detection For Interception By A Humanoid Robot, Saltanat B. Tazhibayeva

Open Access Theses

Interception of a moving object with an autonomous robot is an important problem in robotics. It has various application areas, such as in an industrial setting where products on a conveyor would be picked up by a robotic arm, in the military to halt intruders, in robotic soccer (where the robots try to get to the moving ball and try to block an opponent's attempt to pass the ball), and in other challenging situations. Interception, in and of itself, is a complex task that demands a system with target recognition capability, proper navigation and actuation toward the moving target. There …

Recommending Investors For Crowdfunding Projects, Jisun An, Daniele Quercia, Jon Crowcroft

Research Collection School Of Computing and Information Systems

To bring their innovative ideas to market, those embarking in new ventures have to raise money, and, to do so, they have often resorted to banks and venture capitalists. Nowadays, they have an additional option: that of crowdfunding. The name refers to the idea that funds come from a network of people on the Internet who are passionate about supporting others' projects. One of the most popular crowdfunding sites is Kickstarter. In it, creators post descriptions of their projects and advertise them on social media sites (mainly Twitter), while investors look for projects to support. The most common reason for …

A Hamming Embedding Kernel With Informative Bag-Of-Visual Words For Video Semantic Indexing, Feng Wang, Wen-Lei Zhao, Chong-Wah Ngo, Bernard Merialdo

Research Collection School Of Computing and Information Systems

In this article, we propose a novel Hamming embedding kernel with informative bag-of-visual words to address two main problems existing in traditional BoW approaches for video semantic indexing. First, Hamming embedding is employed to alleviate the information loss caused by SIFT quantization. The Hamming distances between keypoints in the same cell are calculated and integrated into the SVM kernel to better discriminate different image samples. Second, to highlight the concept-specific visual information, we propose to weight the visual words according to their informativeness for detecting specific concepts. We show that our proposed kernels can significantly improve the performance of concept …

Markov Chain Monte Carlo Bayesian Predictive Framework For Artificial Neural Network Committee Modeling And Simulation, Michael S. Goodrich

Computational Modeling & Simulation Engineering Theses & Dissertations

A logical inference method of properly weighting the outputs of an Artificial Neural Network Committee for predictive purposes using Markov Chain Monte Carlo simulation and Bayesian probability is proposed and demonstrated on machine learning data for non-linear regression, binary classification, and 1-of-k classification. Both deterministic and stochastic models are constructed to model the properties of the data. Prediction strategies are compared based on formal Bayesian predictive distribution modeling of the network committee output data and a stochastic estimation method based on the subtraction of determinism from the given data to achieve a stochastic residual using cross validation. Performance for Bayesian …

Stfu Noob!: Predicting Crowdsourced Decisions On Toxic Behavior In Online Games, Jeremy Blackburn, Haewoon Kwak

Research Collection School Of Computing and Information Systems

One problem facing players of competitive games is negative, or toxic, behavior. League of Legends, the largest eSport game, uses a crowdsourcing platform called the Tribunal to judge whether a reported toxic player should be punished or not. The Tribunal is a two stage system requiring reports from those players that directly observe toxic behavior, and human experts that review aggregated reports. While this system has successfully dealt with the vague nature of toxic behavior by majority rules based on many votes, it naturally requires tremendous cost, time, and human efforts. In this paper, we propose a supervised learning approach …

The Use Of The Blackboard Architecture For A Decision Making System For The Control Of Craft With Various Actuator And Movement Capabilities, Jeremy Straub, Hassan Reza

Jeremy Straub

This paper provides an overview of an approach to the control of multiple craft with heterogeneous movement and actuation characteristics that is based on the Blackboard software architecture. An overview of the Blackboard architecture is provided. Then, the operational and mission requirements that dictate the need for autonomous control are characterized and the utility of the Blackboard architecture is for meeting these requirements is discussed. The performance of a best-path solver and naïve solver are compared. The results demonstrate that the best-path solver outperforms the naïve solver in the amount of time taken to generate a solution; however, the number …

Openorbiter Operating Software, Dayln Limesand, Christoffer Korvald, Jeremy Straub, Ronald Marsh

Jeremy Straub

The operating software team of the OpenOrbiter project has been tasked with developing software for general spacecraft maintenance, performing mission tasks and the monitoring of system critical aspects of the spacecraft. To do so, the team is developing an autonomous system that will be able to continuously check sensors for data, and schedule tasks that pertain to the current mission and general maintenance of the onboard systems. Development in support of these objectives is ongoing with work focusing on the completion of the development of a stable system. This poster will present an overview of current work on the project …

Session E-4: Teach Students Robotics!, Pat Patankar

Professional Learning Day

In this session, we will discuss how to introduce your students to Robotics. We will talk about the programing environment, the hardware that is required, and the resources you'll need to get your program up and running. We will discuss how to acquire them. A number of them are free - and others can be obtained by the students themselves (in fact they may already have what they'll need!).

Building Thinc: User Incentivization And Meeting Rescheduling For Energy Savings, Jun Young Kwak, Debarun Kar, William Haskell, Pradeep Reddy Varakantham, Milind Tambe

Research Collection School Of Computing and Information Systems

This paper presents THINC, an agent developed for saving energy in real-world commercial buildings. While previous work has presented techniques for computing energy-efficient schedules, it fails to address two issues, centered on human users, that are essential in real-world agent deployments: (i) incentivizing users for their energy saving activities and (ii) interacting with users to reschedule key “energy-consuming” meetings in a timely fashion, while handling the uncertainty in such interactions. THINC addresses these shortcomings by providing four new major contributions. First, THINC computes fair division of credits from energy savings. For this fair division, THINC provides novel algorithmic advances for …