Index Catalog // ScholarsArchive@OSU

Charting a Path Through an Uncertain Future: Lessons from Our History

Creator:: Spain, Glen
Abstract:: The history of our fishing industry gives us insights into the ‘megatrends’ and socioeconomic forces that will shape our future, and also provides us tools for addressing the problems of the present. The future will be less predictable than the past, due to a wide variety of economic and physical...
Resource Type:: Research Paper
Full Text:: Through an Uncertain Future: Lessons from Our History Glen Spain, Pacific Coast Federation of Fishermen’s

Proposed metrics for transfer learning

Creator:: Oregon State University. Dept. of Computer Science and Dietterich, Thomas Glen
Abstract:: Summary: Four proposed metrics: [1] average relative reduction in training time (sample size, number of training experiences) [2] jumpstart (initial advantage of transfer algorithm) [3] handicap (how long it takes the no-transfer algorithm to overcome the jumpstart) [4] asymptotic advantage (how much better the transfer learning algorithm does in the...
Resource Type:: Research Paper
Full Text:: e r f o r m a n c e learning curve for algorithm B learning curve for algorithm AB p*AB

A POMDP approximation algorithm that anticipates the need to observe

Creator:: Oregon State University. Dept. of Computer Science, Zubek, Valentina Bayer, and Dietterich, Thomas Glen
Abstract:: This paper introduces the even-odd POMDP an approximation to POMDPs Partially Observable Markov Decision Problems in which the world is assumed to be fully observable every other time step. This approximation works well for problems with a delayed need to observe. The even-odd POMDP can be converted into an equivalent...
Resource Type:: Research Paper
Full Text:: method can be expected to solve all POMDPs e�ectively and e ciently� Indeed� the problem of �nding

A POMDP approximation algorithm that anticipates the need to observe

Creator:: Oregon State University. Dept. of Computer Science, Bayer, Valentina, and Dietterich, Thomas Glen
Abstract:: This paper introduces the even-odd POMDP, an approximation to POMDPs in which the world is assumed to be fully observable every other time step. The even-odd POMDP can be converted into an equivalent MDP, the 2MDP, whose value function, V*[subscript 2MDP], can be combined online with a 2-step lookahead search...
Resource Type:: Research Paper
Full Text:: model also means that no single method can be ex- pected to solve all POMDPs e�ectively and e�ciently

Improving SVM accuracy by training on auxiliary data sources

Creator:: Oregon State University. Dept. of Computer Science, Wu, Pengcheng, and Dietterich, Thomas Glen
Abstract:: The standard model of supervised learning assumes that training and test data are drawn from the same underlying distribution. This paper explores an application in which a second, auxiliary, source of data is available drawn from a different distribution. This auxiliary data is more plentiful, but of significantly lower quality,...
Resource Type:: Research Paper
Full Text:: 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 1 2 3 4 5 6 E rr or r at e Number of

Two heuristics for solving POMDPs having a delayed need to observe

Creator:: Oregon State University. Dept. of Computer Science, Zubek, Valentina Bayer, and Dietterich, Thomas Glen
Abstract:: A common heuristic for solving Partially Observable Markov Decision Problems POMDPs is to first solve the underlying Markov Decision Process MDP and then construct a POMDP policy by performing a fixed depth lookahead search in the POMDP and evaluating the leaf nodes using the MDP value function. A problem with...
Resource Type:: Research Paper
Full Text:: prove that both heuristics compute value functions that are upper bounded by �i�e�� bet� ter than� the

Pruning improves heuristic search for cost-sensitive learning

Creator:: Oregon State University. Dept. of Computer Science, Zubek, Valentina Bayer, and Dietterich, Thomas Glen
Abstract:: This paper addresses cost-sensitive classification in the setting where there are costs for measuring each attribute as well as costs for misclassification errors. We show how to formulate this as a Markov Decision Process in which the transition model is learned from the training data. Specifically we assume a set...
Resource Type:: Research Paper
Full Text:: and just predict that everyone is healthy But a false negative diagnosis i e � declaring a sick

Integrating learning from examples into the search for diagnostic policies

Creator:: Oregon State University. Dept. of Computer Science, Bayer-Zubek, Valentina, and Dietterich, Thomas Glen
Abstract:: This paper studies the problem of learning diagnostic policies from training examples. A diagnostic policy is a complete description of the decision-making actions of a diagnostician (i.e., tests followed by a diagnostic decision) for all possible combinations of test results. An optimal diagnostic policy is one that minimizes the expected...
Resource Type:: Research Paper

Machine learning for cognitive networks : technology assessment and research challenges

Creator:: Oregon State University. Dept. of Computer Science, Dietterich, Thomas Glen, and Langley, Pat
Abstract:: The field of machine learning has made major strides over the last 20 years. This document summarizes the major problem formulations that the discipline has studied, then reviews three tasks in cognitive networking and briefly discusses how aspects of those tasks fit these formulations. After this, it discusses challenges for...
Resource Type:: Research Paper

Learning and reasoning

Creator:: Oregon State University. Dept. of Computer Science and Dietterich, Thomas Glen
Abstract:: What is the relationship between learning and reasoning? Much recent work in machine learning has been criticized for focusing on learning and ignoring reasoning. This paper attempts to describe the various ways in which machine learning research has (and has not) incorporated reasoning. The paper argues that there are important...
Resource Type:: Research Paper
Full Text:: issue on constraints in data mining. SIGKDD Explorations, 4 (1). Buchanan, B., & Shortliffe, E. (1984

ScholarsArchive@OSU

Charting a Path Through an Uncertain Future: Lessons from Our History

Proposed metrics for transfer learning

A POMDP approximation algorithm that anticipates the need to observe

A POMDP approximation algorithm that anticipates the need to observe

Improving SVM accuracy by training on auxiliary data sources

Two heuristics for solving POMDPs having a delayed need to observe

Pruning improves heuristic search for cost-sensitive learning

Integrating learning from examples into the search for diagnostic policies

Machine learning for cognitive networks : technology assessment and research challenges

Learning and reasoning

Limit your search

Academic Affiliation

Advisor

Commencement Year

Conference Name

Conference Section/Track

Creator

Contributor

Date

Decade

Degree Field

Degree Level

Degree Name

File Format

Language

License

Non-Academic Affiliation

Peer Reviewed

Resource Type

Rights Statement

Subject

Search Constraints

Search Results

Limit your search