Person:
Rehg, James M.

Permanent Link
https://hdl.handle.net/1853/71627
Associated Organization(s)
Organizational Unit
ORCID
ArchiveSpace Name Record

Filters

2003 - 2009 4
4
4
4
4
Reset filters

Settings
search.filters.applied.f.isOrgUnitOfPublicationTitle: Center for Experimental Research in Computer Systems ×

Publication Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    ITR/SY: a distributed programming infrastructure for integrating smart sensors
    (Georgia Institute of Technology, 2009-11-30) Ramachandran, Umakishore ; DeWeerth, Stephen P. ; Mackenzie, Kenneth M. ; Starner, Thad ; Hutto, Phil ; Wolenetz, Matt ; Rehg, James M.
  • Thumbnail Image
    Item
    Improving the Classification of Software Behaviors using Ensembles
    (Georgia Institute of Technology, 2005) Bowring, James Frederick ; Harrold, Mary Jean ; Rehg, James M.
    One approach to the automatic classification of program behaviors is to view these behaviors as the collection of all the program's executions. Many features of these executions, such as branch profiles, can be measured, and if these features accurately predict behavior, we can build automatic behavior classifiers from them using statistical machine-learning techniques. Two key problems in the development of useful classifiers are (1) the costs of collecting and modeling data and (2) the adaptation of classifiers to new or unknown behaviors. We address the first problem by concentrating on the properties and costs of individual features and the second problem by using the active-learning paradigm. In this paper, we present our technique for modeling a data-flow feature as a stochastic process exhibiting the Markov property. We introduce the novel concept of databins to summarize, as Markov models, the transitions of values for selected variables. We show by empirical studies that databin-based classifiers are effective. We also describe ensembles of classifiers and how they can leverage their components to improve classification rates. We show by empirical studies that ensembles of control-flow and data-flow based classifiers can be more effective than either component classifier.
  • Thumbnail Image
    Item
    State Management in .NET Web Services
    (Georgia Institute of Technology, 2003) Song, Xiang ; Jeong, Namgeun ; Hutto, Phillip W. ; Ramachandran, Umakishore ; Rehg, James M.
    In the paper, we identify a problem for certain applications wishing to use the web service paradigm to enhance interoperability: rapid, robust state maintenance. We classify two kinds of state: application state and session state. While many features are available to support session data, special mechanisms for application state maintenance are less well developed. Application state maintenance is integral to providing reliable, fault-tolerant web services. We discuss three different models to solve the problem and compare the advantages and disadvantages of each. Experimental results show that the choice of which model to use depends on application requirements. Many important emerging applications will involve the communication of potentially large time-sequenced data streams among heterogeneous clients with varying QoS requirements. D-Stampede.NET is an implementation of a system designed to support the development of such applications. We describe our web service entation along with our state server solution to the application state management problem. A simple demo application is described and measured to validate performance.
  • Thumbnail Image
    Item
    Software Behavior: Automatic Classification and its Applications
    (Georgia Institute of Technology, 2003) Bowring, James Frederick ; Rehg, James M. ; Harrold, Mary Jean
    A program's behavior is ultimately the collection of all its executions. This collection is diverse, unpredictable, and generally unbounded. Thus it is especially suited to statistical analysis and machine learning techniques. We explore the thesis that 1st- and 2nd-order Markov models of event-transitions are effective predictors of program behavior. We present a technique that models program executions as Markov models, and a clustering method for Markov models that aggregates multiple program executions, yielding a statistical description of program behaviors. With this approach, we can train classifiers to recognize specific behaviors emitted by an execution without knowledge of inputs or outcomes. We evaluate an application of active learning to the efficient refinement of our classifiers by conducting three empirical studies that explore a scenario illustrating automated test plan augmentation. We present a set of potential research questions and applications that our work suggests.