Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY...

21
Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk & Decision Sciences Group March 7, 2006

description

3 Illuminating the Path Visual Analytics Agenda - Recommendations –Rec. 4.10: Develop new methods and technologies for capturing and representing information quality and uncertainty –Rec. 4.11: Determine the applicability of confidence assessment in the identification, representation, aggregation, and communication of uncertainties in both the information and the analytical methods used in their assessment. – Summary Rec: Develop methods and principles for representing data quality, reliability, and certainty measures throughout the data transformation and analysis process

Transcript of Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY...

Page 1: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

Click to edit Master title style

Click to edit Master subtitle style

A PRACTICAL LOOK AT UNCERTAINTY MODELING

Steve UnwinRisk & Decision Sciences Group

March 7, 2006

Page 2: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

2

"The fundamental cause of trouble in the world today is that the stupid are cock-sure while the intelligent are full

of doubt.“

Bertrand Russell

Page 3: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

3

Illuminating the Path• Visual Analytics Agenda - Recommendations

– Rec. 4.10: Develop new methods and technologies for capturing and representing information quality and uncertainty

– Rec. 4.11: Determine the applicability of confidence assessment in the identification, representation, aggregation, and communication of uncertainties in both the information and the analytical methods used in their assessment.

– Summary Rec: Develop methods and principles for representing data quality, reliability, and certainty measures throughout the data transformation and analysis process

Page 4: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

4

Uncertainty Analysis as Resource to Visual Analytics

• VA Agenda

– Develop new methods and technologies for capturing and representing information quality and uncertainty

– Determine the applicability of confidence assessment in the identification, representation, aggregation, and communication of uncertainties in both the information and the analytical methods used in their assessment.

– Develop methods and principles for representing data quality, reliability, and certainty measures throughout the data transformation and analysis process

• UA Insight

– Probabilistic techniques• Elicitation methods• Aggregation methods• Information-theoretic approaches

– Nonprobabilistic alternatives• Dempster-Shafer• Possibility theory

– Uncertainty propagation techniques• Analytic• Numerical

– Risk communication• Risk representation• Decision-analysis methods

Page 5: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

5

MEASURING UNCERTAINTY

CLASSICALMETHODS BAYESIAN

METHODS

NON-PROBABILISTIC

METHODS

Page 6: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

6

Classical Statistics

• Focus on Aleatory Uncertainty– random variation inherent in the system

• Sampling produces confidence intervals• Need a sampling model

– Generally unavailable for many real-world complex situations

• Confidence intervals are not probability intervals– Propagation difficulties in even the simplest models

Page 7: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

7

Bayesianism

• de Finetti, Ramsey, Savage (1920s-50s)• Subjectivism – Epistemic Probabilities

– Probability as a degree of belief• Classicists are coin tossers• Bayesians are believers

– What is the basis for forming probability?• “ Probabilities do not exist”

– Bruno de Finetti

Page 8: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

8

Problems with Bayesianism• Because probabilities don’t exist, they have to be

created– but how?

• Bayes’ Theorem• Subjectivity is explicit

– judgment of evidence• Do probabilities really reflect the way we conceive

belief?– is probability theory a good theory of evidence?– what are the options?

Page 9: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

9

One Option:Dempster-Shafer Theory

• Withholding belief distinct from disbelief• Seahawks or Steelers will win?• Set of possibilities: {sea, steel}• Probability theory:

– Weight of evidence attached to each exclusive possibility– p(sea), p(steel)

• D-S theory:– Weight of evidence attached to each subset– m(Ø), m(sea), m(steel), m(sea U steel)

• Allows: m(sea U steel) = 1, all other m=0– A compelling ignorance

Page 10: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

10

Support and Plausibility

• Probability replaced by two belief measures:– Each calculated from weights of evidence– bel(sea) is the support for proposition ‘sea’– pl(sea) is the plausibility of ‘sea’– bel(sea) ≤ pl(sea)– Upper and lower “probabilities”

• Complete ignorance• SDU: bel(sea) = 0, pl(sea) = 1, i.e., complete

ignorance on the matter of proposition ‘sea’

Page 11: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

11

Complementary Cumulative Belief Functions

ESD Sensor System On-Demand Failure Rate

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.0E-04 1.0E-03 1.0E-02 1.0E-01 1.0E+00

Failure Rate per Demand

Belie

f Met

ric

Complementary Cumulative Support/ Belief

Complementary Cumulative Plausibility

Complementary Cumulative Probability

Page 12: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

12

Possibility Theory

• Genesis in fuzzy sets• Possibility is an uncertainty measure that

mirrors the fuzzy set notion of imprecision

Page 13: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

13

The Set of Tall Men

0

0.2

0.4

0.6

0.8

1

5' 8" 5' 9" 5' 10" 5' 11" 6' 0" 6' 1" 6' 2"

Height

Mem

bers

hip

to S

et TallVery Tall

m(h)

m'(h) = m2(h)

Page 14: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

14

Possibility Theory

• 2-tier belief: possibility and necessity• nec(X) ≤ pos(X)• Distinctive combinatorial logic

– nec(X^Y) = min[nec(X), nec(Y)]– pos(XvY) = max[pos(X), pos(Y)]

• No conceptual connection to probability– although probability/possibility can co-exist

Page 15: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

15

Possibilistic Interpretationof Intelligence Statements (Heuer)

Probability

Possibility

Chances are slight

Little chanceBetter than even

Highly likely

Page 16: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

16

Experience with Nonprobabilistic Methods

• Not all good:– Standardization of belief metrics?– Treatment of dependences?– Treatment of conflicting evidence?– Computational demands?– Interpretation of results?– Incorporation into decision process?

• Plan B: Confront the problems with probabilistic methods

Page 17: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

17

Principled Basis for Probability Formulation

• Analysts uncomfortable producing probabilities– justified discomfort

• Alternative:– Produce defensible basis for probability formulation based on

nonprobabilistic judgment• Maximize expression of uncertainty subject to judged

constraints• Borrow uncertainty metrics from:

– statistical mechanics– information theory

• Entropy = -∑i pi.ln pi – discrete probability distribution, pi

Page 18: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

18

Application of Information-Theoretic Methods

• Two USNRC programs:– QUEST- SNL

• Quantitative uncertainty evaluation of source terms

– QUASAR – BNL• Quantitative uncertainty analysis of severe accident releases

• Both studies used the same form of input to the same deterministic models– non-probabilistic input

• expert-generated input parameter uncertainty ranges

• QUEST: Bounding analysis

• QUASAR: Information Theory used to generate probability distributions from bounds

• Probabilistic analysis internal to methodology – no elicitation of probability

Page 19: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

19

Information Theory and the Preservation of Uncertainty

Uncertainty Bands

1.00E-04 1.00E-03 1.00E-02 1.00E-01 1.00E+00

QUEST

QUASAR

QUEST

QUASARI-131

Cs-137

Release Fraction

Page 20: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

20

Uncertainty Analysis as Resource to Visual Analytics

• VA Agenda

– Develop new methods and technologies for capturing and representing information quality and uncertainty

– Determine the applicability of confidence assessment in the identification, representation, aggregation, and communication of uncertainties in both the information and the analytical methods used in their assessment.

– Develop methods and principles for representing data quality, reliability, and certainty measures throughout the data transformation and analysis process

• UA Insight

– Probabilistic techniques• Elicitation methods• Aggregation methods• Information-theoretic approaches

– Nonprobabilistic alternatives• Dempster-Shafer• Possibility theory

– Uncertainty propagation techniques• Analytic• Numerical

– Risk communication• Risk representation• Decision-analysis methods

Page 21: Click to edit Master title style Click to edit Master subtitle style A PRACTICAL LOOK AT UNCERTAINTY MODELING Steve Unwin Risk  Decision Sciences Group.

21

Merit Criteriafor Uncertainty Analysis in Intel

• Makes the analyst’s job easier• Represents strength of evidence intuitively• Can reflect dissonant evidence• Appropriately propagates uncertainty from analyst

to decision-maker• Characterizes output uncertainty in a standardized

and interpretable way• Computationally tractable• Promotes insight