UID:
almafu_9960119228402883
Format:
1 online resource (xii, 332 pages) :
,
digital, PDF file(s).
Edition:
1st pbk. ed.
ISBN:
1-139-92717-5
,
1-139-93008-7
,
0-511-84060-8
Content:
The authors explain the ways in which uncertainty is an important factor in the problems of risk and policy analysis. This book outlines the source and nature of uncertainty, discusses techniques for obtaining and using expert judgment, and reviews a variety of simple and advanced methods for analyzing uncertainty.
Note:
Title from publisher's bibliographic system (viewed on 05 Oct 2015).
,
Cover -- Half-title -- Title -- Copyright -- Contents -- Preface -- 1. Introduction -- 1.1. Does Uncertainty Really Matter? -- 1.2. Outline of the Book -- References -- 2. Recent Milestones -- 2.1. Reactor Safety -- 2.2. Air Pollution -- 2.3. Ozone and Chlorofluorocarbons -- References -- 3. An Overview of Quantitative Policy Analysis -- 3.1. Policy Research and Policy Analysis -- 3.2. Policy Research and Analysis versus Natural Science -- 3.2.1. Empirical Testing -- 3.2.2. Documentation and Reproducibility -- 3.2.3. Reporting Uncertainty -- 3.2.4. Peer Review -- 3.2.5. Debate -- 3.3. Which Comes First: Goals or Analysis? -- 3.4. Philosophical Frameworks for Analysis -- 3.4.1. Criteria for Decision Making -- 3.4.2. General Policy Strategies -- 3.4.3. Being Consistent about Criteria and Strategies -- 3.4.4. Choosing Explicitly -- 3.5. Setting the Boundaries -- 3.6. Motivations for Undertaking Policy Analysis -- 3.7. All Motivations Require Some Focus on Substance -- 3.8. Ten Commandments for Good Policy Analysis -- 3.8.1. Do Your Homework with Literature, Experts, and Users -- 3.8.2. Let the Problem Drive the Analysis -- 3.8.3. Make the Analysis as Simple as Possible, but No Simpler -- 3.8.4. Identify All Significant Assumptions -- 3.8.5. Be Explicit about Decision Criteria and Policy Strategies -- 3.8.6. Be Explicit about Uncertainties -- 3.8.7. Perform Systematic Sensitivity and Uncertainty Analysis -- 3.8.8. Iteratively Refine the Problem Statement and the Analysis -- 3.8.9. Document Clearly and Completely -- 3.8.10. Expose to Peer Review -- 3.9. Why Consider Uncertainty? -- References -- 4. The Nature and Sources of Uncertainty -- 4.1. Introduction -- 4.2. The Nature of Probability -- 4.2.1. The Frequentist View -- 4.2.2. The Personalist or Bayesian View -- 4.2.3. The Clarity Test -- 4.3. Types of Quantity -- 4.3.1. Empirical Quantities.
,
4.3.2. Defined Constants -- 4.3.3. Decision Variables -- 4.3.4. Value Parameters -- 4.3.5. Index Variables -- 4.3.6. Model Domain Parameters -- 4.3.7. State Variables -- 4.3.8. Outcome Criteria -- 4.4. Sources of Uncertainty in Empirical Quantities -- 4.4.1. Random Error and Statistical Variation -- 4.4.2. Systematic Error and Subjective Judgment -- 4.4.3. Linguistic Imprecision -- 4.4.4. Variability -- 4.4.5. Randomness and Unpredictability -- 4.4.6. Disagreement -- 4.4.7. Approximations -- 4.5. Uncertainty about Model Form -- 4.6. Inputs, Outputs, and Model Solution Methods -- References -- 5. Probability Distributions and Statistical Estimation -- 5.1. Introduction -- 5.2. Characterizing Probability Distributions -- 5.2.1. Cumulative Distribution Functions -- 5.2.2. Fractiles -- 5.2.3. Moments -- 5.3. Estimation -- 5.3.1. Classical Estimation -- 5.3.2. Estimating Moments -- 5.3.3. Estimating Fractiles -- 5.3.4. Bayesian Estimation -- 5.4. Common and Useful Probability Distribution Functions -- 5.4.1. Normal Distribution -- 5.4.2. Lognormal Distribution -- 5.4.3. Exponential Distribution -- 5.4.4. Poisson Distribution -- 5.4.5. Gamma Distribution -- 5.4.6. Weibull Distribution -- 5.4.7. Uniform Distribution -- 5.4.8. Triangular Distribution -- 5.4.9. Beta Distribution -- 5.4.10. Probability of an Event: The Bernoulli Distribution -- 5.4.11. Binomial Distribution -- 5.5. Multivariate Distributions -- 5.6. Evaluating the Fit of a Distribution -- References -- 6. Human Judgment about and with Uncertainty -- 6.1. The Psychology of Judgment under Uncertainty -- 6.2. The Psychology of Probability Assessment -- 6.3. Evaluating Subjective Probability Judgments -- 6.3.1. Scoring Rules -- 6.3.2. Measuring Calibration -- 6.4. Techniques for Encoding Probabilities -- 6.4.1. Encoding Discrete Probabilities -- 6.4.2. Encoding Continuous Distributions.
,
6.4.3. Disaggregation -- 6.4.4. The Use of Reasons and Disconfirming Information -- 6.4.5. Training in Calibration -- 6.4.6. Covariation and Dependence -- 6.4.7. Multivariate Distributions -- 6.4.8. Aids in Elicitation -- 6.4.9. Is Any Technique "Best"? -- 6.5. Are Experts Different? -- 6.6. Conclusions -- References -- 7. Performing Probability Assessment -- 7.1. The Stanford/SRI Assessment Protocol -- 7.2. A Protocol We Have Used -- 7.3. The Wallsten/EPA Protocol -- 7.4. The Attributes of a "Good" Protocol -- 7.5. Experts Who "Can't or Won't Play the Elicitation Game -- 7.6. Issues of Complexity and Correlation -- 7.7. Multiple Experts with Different Opinions -- 7.8. Limitations, Problems, and Risks -- References -- 8. The Propagation and Analysis of Uncertainty -- 8.1. Introduction -- 8.2. Basic Concepts -- 8.3. Analytic Methods -- 8.3.1. Approximation from the Taylor Series -- 8.3.2. First Order Approximation -- 8.3.3. Weighted Sums -- 8.3.4. Products of Powers with Log Transform -- 8.3.5. Products of Powers with Relative Error -- 8.3.6. Higher Order Approximations -- 8.3.7. Finding Derivatives -- 8.3.8. Moments and Other Properties of Distributions -- 8.3.9. Advantages and Disadvantages of the Analytic Methods -- 8.4. Discrete Distributions and Decision Analysis -- 8.4.1. Discretizing Continuous Distributions -- 8.4.2. The Method of Discrete Probability Distributions -- 8.4.3. Decision Analysis and Uncertainty Analysis -- 8.5. Monte Carlo and Other Sampling Methods -- 8.5.1. Monte Carlo Has Linear Complexity -- 8.5.2. Selecting the Sample Size: Uncertainty about the Mean -- 8.5.3. Estimating Confidence Intervals for Fractiles -- 8.5.4. How Many Runs Are Enough? -- 8.5.5. Variance Reduction Techniques -- 8.5.6. Stratified Sampling and Latin Hypercube Sampling -- 8.5.7. Generating Correlated Input Variables -- 8.5.8. Importance Sampling.
,
8.5.9. Measures of Uncertainty Importance -- 8.6. Fourier Amplitude Sensitivity Test -- 8.7. Response Surface Methods -- 8.8. Selecting a Method -- 8.8.1. Criteria -- 8.8.2. When to Use the Method of Moments -- 8.8.3. Discrete Probability Tree vs. Monte Carlo -- 8.8.4. What Type of Sampling Scheme to Use? -- 8.8.5. When to Use Response Surface Methods -- 8.8.6. The Importance of Software -- References -- 9. The Graphic Communication of Uncertainty -- 9.1. Introduction -- 9.2. Displaying One-dimensional Probability Distributions -- 9.3. Experimental Findings on Communicating Uncertainty about the Value of a Single Uncertain Quantity -- 9.3.1. Experimental Design -- 9.3.2. Results -- 9.3.3. Discussion -- 9.4. Graphing Two-dimensional Uncertainty -- 9.5. The Abstraction Dilemma -- 9.6. Abstraction as Projection -- 9.7. Multidimensional Examples -- 9.8. The Importance of Audience -- References -- 10. Analytical A Software Tool for Uncertainty Analysis and Model Communication -- 10.1. Black Box Models as an Obstacle to Communication -- 10.2. Visual Display of Model Structure with Influence Diagrams -- 10.3. Integrating Documentation with the Model Structure -- 10.4. Organization of a Model as a Hierarchy of Modules -- 10.5. Tools for Modeling and Analysis of Uncertainty -- 10.5.1. Selection and Assessment of Probability Distributions -- 10.5.2. Assessment of Dependent Distributions -- 10.5.3. Display of Probability Distributions -- 10.5.4. Methods of Propagating Uncertainty -- 10.5.5. Uncertainty Analysis -- 10.6. Progressive Refinement and Array Abstraction -- 10.7. Dynamic Models: Influence Diagrams and Systems Dynamics -- 10.8. Collaborative Model Development -- 10.9. Conclusions -- References -- 11. Large and Complex Models -- 11.1. What Are "Large" and "Complex" Models? -- 11.2. Large and Complex Policy Models: Some Examples.
,
11.2.1. General Purpose Regulatory Models -- 11.2.2. Global systems models -- 11.2.3. Insights from Experiences with Global Systems Models -- 11.3. Limits Imposed by Stochastic Processes -- 11.4. Legitimate Reasons for Building Large and Complex Models -- 11.5. Simplifying Large Research Models for Policy Applications -- References -- 12. The Value of Knowing How Little You Know -- 12.1. The Expected Value of Including Uncertainty (EVIU) -- 12.2. The EVIU and the EVPI -- 12.3. The EVIU and Risk Premium -- 12.4. When Ignoring Uncertainty Doesn't Matter -- 12.5. Cubic Error Loss -- 12.6. The Bilinear or "Newsboy" Loss Function -- 12.7. The Catastrophic or "Plane-catching" Problem -- 12.8. What Is the "Best Estimate," Ignoring Uncertainty? -- 12.9. The Duality of Asymmetry in Loss Function and Probability Distribution -- 12.10. The Use of the EVIU -- 12.11. When to Consider Uncertainty and When to Use the EVIU -- 12.12. Summary -- References -- Index.
,
English
Additional Edition:
ISBN 0-521-42744-4
Additional Edition:
ISBN 0-521-36542-2
Language:
English
URL:
https://doi.org/10.1017/CBO9780511840609
Bookmarklink
