检索结果-内蒙古大学图书馆

Adaptive critic learning with fuzzy utility

Adaptive critic learning with fuzzy utility

NAFIPS 2004 - Annual Meeting of the North American Fuzzy Information Processing Society: Fuzzy Sets in the Heart of the Canadian Rockies

作者： Matzner, Shari A. Shannon, Thaddeus T. NW Computational Intelligence Lab. Systems Science Ph.D. Program Portland State University P.O. Box 751 Portland OR 97207 United States

Adaptive Critic methods, which approximate dynamic programming, have been used successfully for solving optimal control problems. The Adaptive Critic learning algorithm optimizes a secondary utility function that is the sum of the present and an future primary utility. The primary utility function measures the instantaneous cost incurred for the last action taken and the resulting state. The motivation for using a fuzzy primary utility function comes from the set of control problems for which there is only a qualitative definition of performance - for example, success or failure. Previous work in applying Adaptive Critic methods to this type of problem showed that a crisp definition of success resulted in solutions that met the control objective, but in an undesirable manner. An appropriate fuzzy utility function, on the other hand, is able to generate the optimal solution. Another motivation for incorporating fuzzy techniques into the utility function is to overcome measurement noise. Measurement noise has a significant adverse effect on the reliability and speed of Adaptive Critic learning;by incorporating fuzzy sets into the utility function, the effect of the noise can be mitigated.

关键词： Fuzzy sets

来源：评论

学校读者我要写书评

暂无评论

Hamilton's rule applied to reciprocal altruism

Hamilton's rule applied to reciprocal altruism

引用

Congress on Evolutionary Computation

作者： J.A. Fletcher M. Zwick Systems Science Ph.D. Program Portland State University Portland OR USA

Reciprocal altruism and inclusive fitness are generally considered alternative mechanisms by which cooperative, altruistic traits may evolve. Here we demonstrate that very general versions of Hamilton's inclusive fitness rule (developed by Queller) can be applied to traditional reciprocal altruism models such as the iterated prisoner's dilemma. In this way we show that both mechanisms rely fundamentally on the same principle - the positive assortment of helping behaviors. We discuss barriers to this unified view, including phenotype/genotype differences and nonadditive fitness (or utility) functions that are typical of reciprocal altruism models. We then demonstrate how Queller's versions of Hamilton's rule remove these obstacles.

关键词： Equations Genetics Frequency Productivity Yarn Predictive models Costs

来源：评论

学校读者我要写书评

暂无评论

Using Design of Experiments, sensitivity analysis, and hybrid simulation to evaluate changes to a software development process: A case study

Using Design of Experiments, sensitivity analysis, and hybri...

引用

作者： Wakeland, Wayne W. Martin, Robert H. Raffo, David Systems Science Ph.D. Program Portland State University P.O. Box 751 Portland OR 97207 United States Software Management Consulting 2017 Maple Terrace West Linn OR 97068 United States School of Business Administration Portland State University P.O. Box 751 Portland OR United States Portland State University P.O. Box 751 Portland OR 97207 United States

Hybrid simulation models combine the high-level project issues of System Dynamics models with the process detail of discrete event simulation models. Hybrid models not only capture the best of both of these simulation paradigms, but they also are able to address new issues that are important in managing complex real-world development projects that neither the System Dynamics nor Discrete Event simulation paradigms are able to address alone. In order to reap the full benefits from a simulation model, a structured approach for analyzing model results is necessary. This article applies Design of Experiments (DOE) and broad range sensitivity analysis (BRSA) to a hybrid system dynamics and discrete event simulation model of a software development process. DOE is used to analyse the interaction effects, such as the degree to which the impact of the process change depends on worker motivation, schedule pressure and other project environmental variables. The sensitivity of the model to parameter changes over a broad range of plausible values is used to analyse the non-linear aspects of the model. The end result is a deeper insight into the conditions under which the process change will succeed, and improved recommendations for process change design and implementation. In this particular study, significant interactions and non-linearities were revealed, supporting the hypothesis that consideration of these complex effects is essential for insightful interpretation of model results and effective decision-making. Copyright © 2004 John Witey & Sons, Ltd.

关键词： Software engineering

来源：评论

学校读者我要写书评

暂无评论

Linguistic Adaptive Critics for tuning fuzzy controllers

Linguistic Adaptive Critics for tuning fuzzy controllers

引用

Conference of the North American Fuzzy Information Processing Society - NAFIPS

作者： T.T. Shannon NW Computational Intelligence Laboratory Systems Science Program Portland State University Portland OR USA

In this paper we demonstrate the use of Linguistic Adaptive Critics (LACs) for the design of fuzzy control systems. This technique is related to Adaptive Critic based Approximate Dynamic programming as developed and applied in the neurocontrol context. We have previously shown how such techniques can be conveniently generalized to fuzzy controllers or neuro-fuzzy hybrids. We demonstrate LACs by designing a temperature controller for a simple water bath system.

关键词： programmable control Adaptive control Fuzzy control Dynamic programming Control systems Los Angeles Council Temperature control Communication system control Computational intelligence Laboratories

来源：评论

学校读者我要写书评

暂无评论

A comparison of system dynamics and agent-based simulation applied to the study of cellular receptor dynamics

A comparison of system dynamics and agent-based simulation a...

引用

The Hawaii International Conference on System sciences, HICSS 2004

作者： Wakeland, Wayne W. Macovsky, Louis M. Gallaher, Edward J. Aktipis, C. Athena Systems Science Ph.D. Program Portland State University United States Dynamic BioSystems LLC Behav. Neurosci. Pharmacol./P. Oregon Health and Science University United States Department of Psychology University of Pennsylvania United States

Cellular receptor dynamics are often analyzed using differential equations, making system dynamics (SD) a candidate methodology. In some cases it may be useful to model the phenomena at the biomolecular level, especially when concentrations and reaction probabilities are low and might lead to unexpected behavior modes. In such cases, agent-based simulation (ABS) may be useful. We show the application of both SD and ABS to simulate non-equilibrium ligand-receptor dynamics over a broad range of concentrations, where the probability of interaction varies from low to very low. Both approaches offer much to the researcher and are complementary. We did not find a clear demarcation indicating when one paradigm or the other would be strongly preferred, although SD is an obvious choice when studying systems at a high level of aggregation and abstraction, and ABS is well suited to studying phenomena at the level of individual receptors and molecules.

关键词： systems analysis

来源：评论

学校读者我要写书评

暂无评论

Adaptive critic learning with fuzzy utility

Adaptive critic learning with fuzzy utility

引用

Conference of the North American Fuzzy Information Processing Society - NAFIPS

作者： S.A. Matzner T.T. Shannon NW Computational Intelligence Laboratory Systems Science Ph.D. Program Portland State University Portland OR USA

Adaptive critic methods, which approximate dynamic programming, have been used successfully for solving optimal control problems. The adaptive critic learning algorithm optimizes a secondary utility function that is the sum of the present and all future primary utility. The primary utility function measures the instantaneous cost incurred for the last action taken and the resulting state. The motivation for using a fuzzy primary utility function comes from the set of control problems for which there is only a qualitative definition of performance - for example, success or failure. Previous work in applying adaptive critic methods to this type of problem showed that a crisp definition of success resulted in solutions that met the control objective, but in an undesirable manner. An appropriate fuzzy utility function, on the other hand, is able to generate the optimal solution. Another motivation for incorporating fuzzy techniques into the utility function is to overcome measurement noise. Measurement noise has a significant adverse effect on the reliability and speed of adaptive critic learning; by incorporating fuzzy sets into the utility function, the effect of the noise can be mitigated.

关键词： Dynamic programming Optimal control programmable control Adaptive control Cost function Control systems Computational intelligence Fuzzy sets Noise measurement Neural networks

来源：评论

学校读者我要写书评

暂无评论

Improved process monitoring with Independent Components

Improved process monitoring with Independent Components

引用

IEEE/SEMI Conference and Workshop on Advanced Semiconductor Manufacturing

作者： T.T. Shannon D. Abercrombie J. McNames B. Whitefield Systems Science Program Portland State University Portland OR USA LSI Logic Inc. Gresham OR USA IC Design & Test Laboratory Portland State University Portland OR USA

We propose a method for process monitoring of a semiconductor manufacturing process. Independent Component Analysis (ICA) is applied to characterize E-test parameter data. We calculate angular confidence intervals for the model, eliminate marginally significant components, and implement control charts for significant components of interest. Alarms are generated off of deviations in the charted components. Alarms are easily used in process diagnosis based on the interpretation of the independent components.

关键词： Monitoring Independent component analysis Manufacturing processes Testing Principal component analysis Process control Fault diagnosis Large scale integration Microelectronics Semiconductor device manufacture

来源：评论

学校读者我要写书评

暂无评论

Using design of experiments, sensitivity analysis, and hybrid simulation to evaluate changes to a software development process: a case study

引用

Software Process: Improvement and Practice 2004年第2期9卷

作者： Wayne W. Wakeland Robert H. Martin David Raffo Systems Science Ph.D. Program Portland State University P.O. Box 751 Portland OR 97207 USA Software Management Consulting 2017 Maple Terrace West Linn OR 97068 USA School of Business Administration Portland State University P.O. Box 751 Portland Oregon USA

关键词： software process model hybrid simulation design of experiments sensitivity analysis

来源：评论

学校读者我要写书评

暂无评论

Artificial Neural Network Implementation Using Many-Valued Quantum Computing

Artificial Neural Network Implementation Using Many-Valued Q...

引用

International Joint Conference on Neural Networks 2003

作者： Al-Rabadi, Anas N. Lendaris, George G. Systems Science Ph.D. Program Portland State University Portland OR United States

Neural network (NN) implementation using the general scheme of many-valued (MV) quantum computing (QC) is presented in this paper. The proposed method uses the many-valued orthonormal computational basis states to implement such computations. Physical implementation of NNQC is performed by controlling the potential to yield specific wavefunction as a result of solving Shrodinger's equation that governs the dynamics of QC.

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

Modeling CMOS gate characteristics using independent component analysis

Modeling CMOS gate characteristics using independent compone...

引用

International Joint Conference on Neural Networks 2003

作者： Shannon, Thaddeus T. Abercrombie, David McNames, James Systems Science Program Portland State University Portland OR United States LSI Logic Gresham OR United States Electrical Engineering Department Portland State University Portland OR United States

Independent Component Analysis, a technique for blind source separation, is applied to E-test parameter data from a CMOS manufacturing process. Models of variance for n channel transistors are developed using the assumption that discrete steps in the process introduce statistically independent variance components through particular physical mechanisms. We calculate confidence intervals for the models, eliminate marginally significant components and chart the percentage of parameter variance explained by each significant component. The isolated variance sources are easily interpretable based on device physics.

关键词： CMOS integrated circuits

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：