Search Constraints

Start Over
Filtering by: Creator Pfleeger, Thomas G. Remove constraint Creator: Pfleeger, Thomas G. Academic Affiliation Computer Science Remove constraint Academic Affiliation: Computer Science

Search Results

  • Learning at the knowledge level

    Creator:
    Dietterich, Thomas G.
    Abstract:
    When Newell introduced the concept of the knowledge level as a useful level of description for computer systems, he focused on the representation of knowl­edge. This paper applies the knowledge level notion to the problem of knowl­edge acquisition. Two interesting issues arise. First, some existing machine learning programs appear to...
    Resource Type:
    Technical Report
    Full Text:
    Knowledge Level Thomas G. Dietterich Department of Computer Science Oregon State University Corvallis

  • A knowledge level analysis of learning programs

    Creator:
    Dietterich, Thomas G.
    Abstract:
    This chapter develops a taxonomy of learning methods using techniques based on Newell’s knowledge level. Two properties of each system are defined: knowl­edge level predictability and knowledge level learning. A system is predictable at the knowledge level if the principle of rationality can be applied to predict its behavior. A...
    Resource Type:
    Technical Report
    Full Text:
    ANALYSIS OF LEARNING PROGRAMS1 Thomas G. Dietterich Department of Computer Science Oregon State

  • Editorial : exploratory research in machine learning

    Creator:
    Dietterich, Thomas G.
    Abstract:
    Exploratory research contributes to the continued vitality of every discipline. The aim of exploratory research is to identify new tasks-tasks that cannot be solved by existing methods. Once a new task has been found, exploratory research seeks to develop a precise definition of the task and to understand the factors...
    Resource Type:
    Technical Report
    Full Text:
    . Thomas G. Dietterich Editor Department of Computer Science Oregon State University References

  • FORLOG : a logic-based architecture for design

    Creator:
    Dietterich, Thomas G. and Ullman, David G.
    Abstract:
    It is difficult to build intelligent computer-aided design (ICAD) programs using available expert system shells and AI programming languages. To build ICAD programs, tools are needed that support (a) generative search of design spaces, (b) deep search of design spaces to evaluate alternative designs, (c) simultane­ous exploration of alternative designs...
    Resource Type:
    Technical Report
    Full Text:
    Architecture for Design Thomas G. Dietterich Department of Computer Science Oregon State University

  • Artificial intelligence at OSU

    Creator:
    Dietterich, Thomas G. and D'Ambrosio, Bruce
    Abstract:
    This report describes of current research in Artificial Intelligence at Oregon State University. The five areas of active research are ( a) intelligent aids for me­chanical engineering design, (b) active experimentation as a method in machine learning, ( c) techniques for combining logic programming and assumption-based truth maintenance, ( d)...
    Resource Type:
    Technical Report
    Full Text:
    at OSU Thomas G. Dietterich Bruce D'Ambrosia Department of Computer Science Oregon State

  • On learning more concepts

    Creator:
    Almuallim, Hussein and Dietterich, Thomas G.
    Abstract:
    The coverage of a learning algorithm is the number of concepts that can be learned by that algorithm from samples of a given size. This paper asks whether good learning algorithms can be designed by maximizing their coverage. The paper extends a previous upper bound on the coverage of any...
    Resource Type:
    Technical Report
    Full Text:
    Concepts Hussein Almuallim Thomas G. Oietterich Department of Computer Science Oregon State

  • Learning optimal search control for engineering optimization

    Creator:
    Cerbone, Giuseppe and Dietterich, Thomas G.
    Abstract:
    This paper applies learning techniques to make engineering optimization more efficient and reliable. When the function to be optimized is highly non-linear, the search space generally forms several disjoint convex regions . Unless gradient-descent search is begun in the right region, the solution found will be suboptimal. This paper formalizes...
    Resource Type:
    Technical Report
    Full Text:
    Engineering Optimization Giuseppe Cerbone Thomas G. Dietterich Oregon State University Computer Science

  • Error-correcting output codes : a general method for improving multiclass inductive learning programs

    Creator:
    Dietterich, Thomas G. and Bakiri, Ghulum
    Abstract:
    Multiclass learning problems involve finding a definition for an unknown function f(x) whose range is a discrete set containing k > 2 values (i.e., k "classes") . The definition is acquired by studying large collections of training examples of the form (xi, f(xi)) . Existing approaches to this problem include...
    Resource Type:
    Technical Report
    Full Text:
    General Method for Improving Multiclass Inductive Learning Programs Thomas G. Dietterich Ghulum

  • Inductive and numerical methods in knowledge compilation

    Creator:
    Cerbone, Giuseppe and Dietterich, Thomas G.
    Abstract:
    Many important application problems can be formalized as constrained non-linear optimization tasks. However, numerical methods for solving such problems are brittle and do not scale well. Furthermore, they do not provide much insight into the structure of the problem space. This paper describes a method for discovering efficient rules that...
    Resource Type:
    Technical Report
    Full Text:
    Methods in Knowledge Comp\lation Giuseppe Cerbone Thomas G. Dietterich Computer Science Department

  • Efficient algorithms for identifying relevant features

    Creator:
    Almuallim, Hussein and Dietterich, Thomas G.
    Abstract:
    This paper describes efficient methods for exact and approximate implementation of the MIN-FEATURES bias, which prefers consistent hypotheses definable over as few features as possible. This bias is useful for learning domains where many irrelevant features are present in the training data. We first introduce FOCUS-2, a new algorithm that...
    Resource Type:
    Technical Report
    Full Text:
    for Identifying Relevant Features Hussein Almuallim Thomas G. Oietterich Department of Computer