Virtual environments and simulations are being used increasingly to both visualize and understand data as well as to create scenarios for training and analysis purposes. In this paper, we are interested in the use of simulation and visualization of interactive virtual agents to create realistic motions for training scenarios. We...
Linear transformation for dimension reduction is a well established problem in the field of machine learning. Due to the numerous observability of parameters and data, processing of the data in its raw form is computationally complex and difficult to visualize. Dimension reduction by means of feature extraction offers a strong...
The results of a machine learning from user behavior can be thought of as a program, and like all programs, it may need to be debugged. Providing ways for the user to debug it matters because without the ability to fix errors, users may find that the learned program’s errors...
This thesis presents a progression of novel planning algorithms that culminates in a new family of diverse Monte-Carlo methods for probabilistic planning domains. We provide a proof for performance guarantees and analyze how these algorithms can resolve some of the shortcomings of traditional probabilistic planning methods. The direct policy search...
End users' programs are fraught with errors, costing companies millions of dollars. One reason may be that researchers and tool designers have not yet focused on end-user debugging strategies. To investigate this possibility, this dissertation presents eight empirical studies and a new strategy-based end-user debugging tool for Excel, called StratCel....
Reinforcement learning in real-world domains suffers from three curses of dimensionality: explosions in state and action spaces, and high
stochasticity or "outcome space" explosion. Multiagent domains are particularly susceptible to these problems. This thesis describes ways to mitigate these curses in several different multiagent domains, including real-time delivery of products...
Communicating dynamic motion content, such as exercise, with a static medium, such as paper, is difficult. The technology exists for presenting 3D animated exercise content to patients; however, the tools for allowing exercise domain experts to effectively author the content do not exist. We conducted two formative studies with exercise...
We took the back-propagation algorithms of Werbos for recurrent and feed-forward neural networks and implemented them on machines with graphics processing units (GPU). The parallelism of these units gave our implementations a 10 to 100 fold increase in speed. For nets with less than 20 neurons the machine performed faster...
This dissertation explores the idea of applying machine learning technologies to help computer users find information and better organize electronic resources, by presenting the research work conducted in the following three applications: FolderPredictor, Stacking Recommendation Engines, and Integrating Learning and Reasoning.
FolderPredictor is an intelligent desktop software tool that helps...
Sequential supervised learning problems arise in many real applications. This dissertation focuses on two important research directions in sequential supervised learning: efficient training and feature induction.
In the direction of efficient training, we study the training of conditional random fields (CRFs), which provide a flexible and powerful model for sequential...
This project aims at implementing Indexing for Web 2.0 Applications. Ajax applications consist of a set of states which are generated by the user through actions such as click, focus, blur etc. events. By saving these DOM states we can index information obtained from dynamically generated web content. To prevent...
Analysis, visualization, and design of vector fields on surfaces have a wide variety of major applications in both scientific visualization and computer graphics. On the one hand, analysis and visualization of vector fields provide critical insights to the flow data produced from simulation or experiments of various engineering processes. On...
Programmers spend a substantial fraction of their debugging time by navigating
through source code, yet little is known about how programmers navigate. With the
continuing growth in size and complexity of software, this fraction of time is likely to
increase, which presents challenges to those seeking both to understand and...
The problem of document classification has been widely studied in machine learning and data mining. In document classification, most of the popular algorithms are based on the bag-of-words representation. Due to the high dimensionality of the bag-of-words representation, significant research has been conducted to reduce the dimensionality via different approaches....
In the field of Human-Computer Interaction, provenance refers to the complete history and genealogy of a document. Provenance can be useful in identifying related resources, such as different versions of the same document or resources used in the creation of a new document. Though methods of provenance collection and applications...
An interdisciplinary study into the theory of design decisions has yielded a model for tracking design changes in hardware/software systems, but it still needs to be applied to a larger system to test its efficiency at tracking important data. This thesis creates an implementation of PLEXIL, a language in development...
The purpose of this project is to load test, and fine tune the loan search functionality of the Broker Blueprint web application, an innovative Business-to-Business (B2B) online service aiding mortgage lenders and brokers in today's highly competitive mortgage market.
Broker Blueprint enables brokers to search for suitable mortgage loans across...
Interconnection networks play important roles in designing high performance computers. Recently two new classes of interconnection networks based on the concept of Gaussian and Eisenstein-Jacobi integers were introduced. In this research, efficient routing and broadcasting algorithms for these networks are developed. Furthermore, constructing edge disjoint Hamiltonian cycles in Gaussian networks...
A financial processor is the most important component of a credit union‘s IT infrastructure. A database storing member demographic information, account balances, and transaction history, it performs financial calculations, such as interest, dividends, and maturities. It also provides a user interface, allowing tellers and financial service representatives to manage accounts...
This paper examines how six online multiclass text classification algorithms perform in the domain of email tagging within the TaskTracer system. TaskTracer is a project-oriented user interface for the desktop knowledge worker. TaskTracer attempts to tag all documents, web pages, and email messages with the projects to which they are...