| |
| |
Foreword | |
| |
| |
Preface | |
| |
| |
Acknowledgments | |
| |
| |
Authors | |
| |
| |
| |
There Is More to Assessing Risk Than Statistics | |
| |
| |
| |
Introduction | |
| |
| |
| |
Predicting Economic Growth: The Normal Distribution and Its Limitations | |
| |
| |
| |
Patterns and Randomness: From School League Tables to Siegfried and Roy | |
| |
| |
| |
Dubious Relationships: Why You Should Be Very Wary of Correlations and Their Significance Values | |
| |
| |
| |
Spurious Correlations: How You Can Always Find a Silly 'Cause' of Exam Success | |
| |
| |
| |
The Danger of Regression: Looking Back When You Need to Look Forward | |
| |
| |
| |
The Danger of Averages | |
| |
| |
| |
What Type of Average? | |
| |
| |
| |
When Averages Alone Will Never Be Sufficient for Decision Making | |
| |
| |
| |
When Simpson's Paradox Becomes More Worrisome | |
| |
| |
| |
Uncertain Information and Incomplete Information: Do Not Assume They Are Different | |
| |
| |
| |
Do Not Trust Anybody (Even Experts) to Properly Reason about Probabilities | |
| |
| |
| |
Chapter Summary | |
| |
| |
Further Reading | |
| |
| |
| |
The Need for Causal, Explanatory Models in Risk Assessment | |
| |
| |
| |
Introduction | |
| |
| |
| |
Are You More Likely to Die in an Automobile Crash When the Weather Is Good Compared to Bad? | |
| |
| |
| |
When Ideology and Causation Collide | |
| |
| |
| |
The Limitations of Common Approaches to Risk Assessment | |
| |
| |
| |
Measuring Armageddon and Other Risks | |
| |
| |
| |
Risks and Opportunities | |
| |
| |
| |
Risk Registers and Heat Maps | |
| |
| |
| |
Thinking about Risk Using Causal Analysis | |
| |
| |
| |
Applying the Causal Framework to Armageddon | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
Measuring Uncertainty: The Inevitability of Subjectivity | |
| |
| |
| |
Introduction | |
| |
| |
| |
Experiments, Outcomes, and Events | |
| |
| |
| |
Multiple Experiments | |
| |
| |
| |
Joint Experiments | |
| |
| |
| |
Joint Events and Marginalization | |
| |
| |
| |
Frequentist versus Subjective View of Uncertainty | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
The Basics of Probability | |
| |
| |
| |
Introduction | |
| |
| |
| |
Some Observations Leading to Axioms and Theorems of Probability | |
| |
| |
| |
Probability Distributions | |
| |
| |
| |
Probability Distributions with Infinite Outcomes | |
| |
| |
| |
Joint Probability Distributions and Probability of Marginalized Events | |
| |
| |
| |
Dealing with More than Two Variables | |
| |
| |
| |
Independent Events and Conditional Probability | |
| |
| |
| |
Binomial Distribution | |
| |
| |
| |
Using Simple Probability Theory to Solve Earlier Problems and Explain Widespread Misunderstandings | |
| |
| |
| |
The Birthday Problem | |
| |
| |
| |
The Monty Hall Problem | |
| |
| |
| |
When Incredible Events Are Really Mundane | |
| |
| |
| |
When Mundane Events Really Are Quite Incredible | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
Bayes' Theorem and Conditional Probability | |
| |
| |
| |
Introduction | |
| |
| |
| |
All Probabilities Are Conditional | |
| |
| |
| |
Bayes' Theorem | |
| |
| |
| |
Using Bayes' Theorem to Debunk Some Probability Fallacies | |
| |
| |
| |
Traditional Statistical Hypothesis Testing | |
| |
| |
| |
The Prosecutor Fallacy Revisited | |
| |
| |
| |
The Defendant's Fallacy | |
| |
| |
| |
Odds Form of Bayes and the Likelihood Ratio | |
| |
| |
| |
Second-Order Probability | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
From Bayes' Theorem to Bayesian Networks | |
| |
| |
| |
Introduction | |
| |
| |
| |
A Very Simple Risk Assessment Problem | |
| |
| |
| |
Accounting for Multiple Causes (and Effects) | |
| |
| |
| |
Using Propagation to Make Special Types of Reasoning Possible | |
| |
| |
| |
The Crucial Independence Assumptions | |
| |
| |
| |
Structural Properties of BNs | |
| |
| |
| |
Serial Connection: Causal and Evidential Trials | |
| |
| |
| |
Diverging Connection: Common Cause | |
| |
| |
| |
Converging Connection: Common Effect | |
| |
| |
| |
Determining Whether Any Two Nodes in a BN Are Dependent | |
| |
| |
| |
Propagation in Bayesian Networks | |
| |
| |
| |
Using BNs to Explain Apparent Paradoxes | |
| |
| |
| |
Revisiting the Monty Hall Problem | |
| |
| |
| |
Simple Solution | |
| |
| |
| |
Complex Solution | |
| |
| |
| |
Revisiting Simpson's Paradox | |
| |
| |
| |
Steps in Building and Running a BN Model | |
| |
| |
| |
Building a BN Model | |
| |
| |
| |
Running a BN Model | |
| |
| |
| |
Inconsistent Evidence | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
Theoretical Underpinnings | |
| |
| |
BN Applications | |
| |
| |
Nature and Theory of Causality | |
| |
| |
Uncertain Evidence (Soft and Virtual) | |
| |
| |
| |
Defining the Structure of Bayesian Networks | |
| |
| |
| |
Introduction | |
| |
| |
| |
Causal Inference and Choosing the Correct Edge Direction | |
| |
| |
| |
The Idioms | |
| |
| |
| |
The Cause-Consequence Idiom | |
| |
| |
| |
Measurement Idiom | |
| |
| |
| |
Definitional/Synthesis Idiom | |
| |
| |
| |
Case 1: Definitional Relationship between Variables | |
| |
| |
| |
Case 2: Hierarchical Definitions | |
| |
| |
| |
Case 3: Combining Different Nodes Together to Reduce Effects of Combinatorial Explosion ("Divorcing") | |
| |
| |
| |
Induction Idiom | |
| |
| |
| |
The Problems of Asymmetry and How to Tackle Them | |
| |
| |
| |
Impossible Paths | |
| |
| |
| |
Mutually Exclusive Paths | |
| |
| |
| |
Distinct Causal Pathways | |
| |
| |
| |
Taxonomic Classification | |
| |
| |
| |
Multiobject Bayesian Network Models | |
| |
| |
| |
The Missing Variable Fallacy | |
| |
| |
| |
Conclusions | |
| |
| |
Further Reading | |
| |
| |
| |
Building and Eliciting Node Probability Tables | |
| |
| |
| |
Introduction | |
| |
| |
| |
Factorial Growth in the Size of Probability Tables | |
| |
| |
| |
Labeled Nodes and Comparative Expressions | |
| |
| |
| |
Boolean Nodes and Functions | |
| |
| |
| |
The Asia Model | |
| |
| |
| |
The OR Function for Boolean Nodes | |
| |
| |
| |
The AND Function for Boolean Nodes | |
| |
| |
| |
M from N Operator | |
| |
| |
| |
NoisyOR Function for Boolean Nodes | |
| |
| |
| |
Weighted Averages | |
| |
| |
| |
Ranked Nodes | |
| |
| |
| |
Background | |
| |
| |
| |
Solution: Ranked Nodes with the TNormal Distribution | |
| |
| |
| |
Alternative Weighted Functions for Ranked Nodes | |
| |
| |
| |
Hints and Tips When Working with Ranked Nodes and NPTs | |
| |
| |
| |
Tip 1: Use the Weighted Functions as Far as Possible | |
| |
| |
| |
Tip 2: Make Use of the Fact That a Ranked Node Parent Has an Underlying Numerical Scale | |
| |
| |
| |
Tip 3: Do Not Forget the Importance of the Variance in the TNormal Distribution | |
| |
| |
| |
Tip 4: Change the Granularity of a Ranked Scale without Having to Make Any Other Changes | |
| |
| |
| |
Tip 5: Do Not Create Large, Deep, Hierarchies Consisting of Rank Nodes | |
| |
| |
| |
Elicitation | |
| |
| |
| |
Elicitation Protocols and Cognitive Biases | |
| |
| |
| |
Scoring Rules and Validation | |
| |
| |
| |
Sensitivity Analysis | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
Numeric Variables and Continuous Distribution Functions | |
| |
| |
| |
Introduction | |
| |
| |
| |
Some Theory on Functions and Continuous Distributions | |
| |
| |
| |
Static Discretization | |
| |
| |
| |
Dynamic Discretization | |
| |
| |
| |
Using Dynamic Discretization | |
| |
| |
| |
Prediction Using Dynamic Discretization | |
| |
| |
| |
Conditioning on Discrete Evidence | |
| |
| |
| |
Parameter Learning (Induction) Using Dynamic Discretization | |
| |
| |
| |
Classical versus Bayesian Modeling | |
| |
| |
| |
Bayesian Hierarchical Model Using Beta-Binomial | |
| |
| |
| |
Avoiding Common Problems When Using Numeric Nodes | |
| |
| |
| |
Unintentional Negative Values in a Node's State Range | |
| |
| |
| |
Potential Division by Zero | |
| |
| |
| |
Using Unbounded Distributions on a Bounded Range | |
| |
| |
| |
Observations with Very Low Probability | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
Hypothesis Testing and Confidence Intervals | |
| |
| |
| |
Introduction | |
| |
| |
| |
Hypothesis Testing | |
| |
| |
| |
Bayes Factors | |
| |
| |
| |
Testing for Hypothetical Differences | |
| |
| |
| |
Comparing Bayesian and Classical Hypothesis Testing | |
| |
| |
| |
Model Comparison: Choosing the Best Predictive Model | |
| |
| |
| |
Accommodating Expert Judgments about Hypotheses | |
| |
| |
| |
Distribution Fitting as Hypothesis Testing | |
| |
| |
| |
Bayesian Model Comparison and Complex Causal Hypotheses | |
| |
| |
| |
Confidence Intervals | |
| |
| |
| |
The Fallacy of Frequentist Confident Intervals | |
| |
| |
| |
The Bayesian Alternative to Confidence Intervals | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
Modeling Operational Risk | |
| |
| |
| |
Introduction | |
| |
| |
| |
The Swiss Cheese Model for Rare Catastrophic Events | |
| |
| |
| |
Bow Ties and Hazards | |
| |
| |
| |
Fault Tree Analysis (FTA) | |
| |
| |
| |
Event Tree Analysis (ETA) | |
| |
| |
| |
Soft Systems, Causal Models, and Risk Arguments | |
| |
| |
| |
KUUUB Factors | |
| |
| |
| |
Operational Risk in Finance | |
| |
| |
| |
Modeling the Operational Loss Generation Process | |
| |
| |
| |
Scenarios and Stress Testing | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
Systems Reliability Modeling | |
| |
| |
| |
Introduction | |
| |
| |
| |
Probability of Failure on Demand for Discrete Use Systems | |
| |
| |
| |
Time to Failure for Continuous Use Systems | |
| |
| |
| |
System Failure Diagnosis and Dynamic Bayesian Networks | |
| |
| |
| |
Dynamic Fault Trees (DFTs) | |
| |
| |
| |
Software Defect Prediction | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
Bayes and the Law | |
| |
| |
| |
Introduction | |
| |
| |
| |
The Case for Bayesian Reasoning about Legal Evidence | |
| |
| |
| |
Building Legal Arguments Using Idioms | |
| |
| |
| |
The Evidence Idiom | |
| |
| |
| |
The Evidence Accuracy Idiom | |
| |
| |
| |
Idioms to Deal with the Key Notions of "Motive" and "Opportunity" | |
| |
| |
| |
Idiom for Modeling Dependency between Different Pieces of Evidence | |
| |
| |
| |
Alibi Evidence Idiom | |
| |
| |
| |
Explaining away Idiom | |
| |
| |
| |
Putting it All Together: Vole Example | |
| |
| |
| |
Using BNs to Expose Further Fallacies of Legal Reasoning | |
| |
| |
| |
The Jury Observation Fallacy | |
| |
| |
| |
The "Crimewatch UK" Fallacy | |
| |
| |
| |
Summary | |
| |
| |
Further Reading | |
| |
| |
| |
The Basics of Counting | |
| |
| |
| |
The Algebra of Node Probability Tables | |
| |
| |
| |
Junction Tree Algorithm | |
| |
| |
| |
Dynamic Discretization | |
| |
| |
| |
Statistical Distributions | |
| |
| |
Index | |