Chapter 16: Making Simple Decision March 23, 2004.

Slides:



Advertisements
Similar presentations
Utility theory U: O-> R (utility maps from outcomes to a real number) represents preferences over outcomes ~ means indifference We need a way to talk about.
Advertisements

Making Simple Decisions Chapter 16 Some material borrowed from Jean-Claude Latombe and Daphne Koller by way of Marie desJadines,
Making Simple Decisions
DSC 3120 Generalized Modeling Techniques with Applications
Utility Axioms Axiom: something obvious, cannot be proven Utility axioms (rules for clear thinking)
CHAPTER 14 Utility Axioms Paradoxes & Implications.

Decision Theory: Single Stage Decisions Computer Science cpsc322, Lecture 33 (Textbook Chpt 9.2) March, 30, 2009.
Cooperating Intelligent Systems Utility theory Chapter 16, AIMA.
91.420/543: Artificial Intelligence UMass Lowell CS – Fall 2010 Lecture 16: MEU / Utilities Oct 8, 2010 A subset of Lecture 8 slides of Dan Klein – UC.
PREFERENCES AND UTILITY
Making Simple Decisions Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 16.
PSY 5018H: Math Models Hum Behavior, Prof. Paul Schrater, Spring 2005 Normative Decision Theory A prescriptive theory for how decisions should be made.
Expected Utility  We discuss “expected utility” in the context of simple lotteries  A generic lottery is denoted (x, y, pr)  The lottery offers payoff.
Decision making. ? Blaise Pascal Probability in games of chance How much should I bet on ’20’? E[gain] = Σgain(x) Pr(x)
Decision Making Under Uncertainty Russell and Norvig: ch 16 CMSC421 – Fall 2006.
Lectures in Microeconomics-Charles W. Upton The Theory of Choice.
Decision. Decision Summary Search for alternatives Descriptive, normative, prescriptive Expected Utility: normative theory of decision Psychology of decision.
CMSC 671 Fall 2003 Class #26 – Wednesday, November 26 Russell & Norvig 16.1 – 16.5 Some material borrowed from Jean-Claude Latombe and Daphne Koller by.
Health Economics & Policy 3 rd Edition James W. Henderson Chapter 4 Economic Evaluation in Health Care.
8-1 CHAPTER 8 Decision Analysis. 8-2 LEARNING OBJECTIVES 1.List the steps of the decision-making process and describe the different types of decision-making.
Intelligent Environments1 Computer Science and Engineering University of Texas at Arlington.
PGM 2003/04 Tirgul7 Foundations of Decision Theory (mostly from Pearl)
Expectimax Evaluation
Decision making Making decisions Optimal decisions Violations of rationality.
Reminder Midterm Mar 7 Project 2 deadline Mar 18 midnight in-class
Phil 148 Choices. Choice Theory: The relationship between probability and action is often complex, however we can use simple mathematical operations (so.
Making Simple Decisions
Axioms Let W be statements known to be true in a domain An axiom is a rule presumed to be true An axiomatic set is a collection of axioms Given an axiomatic.
Advanced AI Rob Lass July 26, Administrative WebCT Problems? Why aren’t homeworks graded!? Midterm next Wed (Aug. 2nd)
Making Simple Decisions Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 16.
Daphne Koller Decision Making Utility Functions Probabilistic Graphical Models Acting.
Chapter 13 February 19, Acting Under Uncertainty Rational Decision – Depends on the relative importance of the goals and the likelihood of.
Quantitative Decision Techniques 13/04/2009 Decision Trees and Utility Theory.
Lecture 3 on Individual Optimization Uncertainty Up until now we have been treating bidders as expected wealth maximizers, and in that way treating their.
CS 188: Artificial Intelligence Spring 2007 Lecture 21:Reinforcement Learning: I Utilities and Simple decisions 4/10/2007 Srini Narayanan – ICSI and UC.
Decision Making Under Uncertainty CMSC 671 – Fall 2010 R&N, Chapters , , material from Lise Getoor, Jean-Claude Latombe, and.
Making Simple Decisions Utility Theory MultiAttribute Utility Functions Decision Networks The Value of Information Summary.
Axiomatic Theory of Probabilistic Decision Making under Risk Pavlo R. Blavatskyy University of Zurich April 21st, 2007.
Decision Making Under Uncertainty CMSC 471 – Spring 2014 Class #12– Thursday, March 6 R&N, Chapters , material from Lise Getoor, Jean-Claude.
CS 188: Artificial Intelligence Fall 2008 Lecture 7: Expectimax Search 9/18/2008 Dan Klein – UC Berkeley Many slides over the course adapted from either.
CS 188: Artificial Intelligence Fall 2007 Lecture 8: Expectimax Search 9/20/2007 Dan Klein – UC Berkeley Many slides over the course adapted from either.
Web-Mining Agents Agents and Rational Behavior Decision-Making under Uncertainty Simple Decisions Ralf Möller Universität zu Lübeck Institut für Informationssysteme.
Making Simple Decisions Chapter 16 Some material borrowed from Jean-Claude Latombe and Daphne Koller by way of Marie desJadines,
Chapter 16 March 25, Probability Theory: What an agent should believe based on the evidence Utility Theory: What the agent wants Decision Theory:
CS 188: Artificial Intelligence Fall 2006 Lecture 8: Expectimax Search 9/21/2006 Dan Klein – UC Berkeley Many slides over the course adapted from either.
Decision Making ECE457 Applied Artificial Intelligence Spring 2007 Lecture #10.
Risk Efficiency Criteria Lecture XV. Expected Utility Versus Risk Efficiency In this course, we started with the precept that individual’s choose between.
Does risk aversion give us a good reason to diversify our charitable portfolio? James Snowden.
Behavioral Finance Preferences Part I Feb 16 Behavioral Finance Economics 437.
The Theory of Rational Choice
CS 188: Artificial Intelligence Fall 2008 Lecture 8: MEU / Utilities 9/23/2008 Dan Klein – UC Berkeley Many slides over the course adapted from either.
CS 188: Artificial Intelligence Spring 2006
Utilities and Decision Theory
Ralf Möller Universität zu Lübeck Institut für Informationssysteme
Making Simple Decisions
ECE457 Applied Artificial Intelligence Fall 2007 Lecture #10
ECE457 Applied Artificial Intelligence Spring 2008 Lecture #10
The Theory of Rational Choice
CS 4/527: Artificial Intelligence
Risk Chapter 11.
Statistical Decision Theory
Rational Decisions and
13. Acting under Uncertainty Wolfram Burgard and Bernhard Nebel
Behavioral Finance Economics 437.
Making Simple Decisions
Making Simple Decisions
Utilities and Decision Theory
Utilities and Decision Theory
Making Simple Decisions
Presentation transcript:

Chapter 16: Making Simple Decision March 23, 2004

16.1 Combining Belief and Desires Under Uncertainty Utility: captures the desirability of a state, U(S) A: action E: evidence EU: expected utility EU(A|E) = ∑ P(Result i (A) | Do(A), E) * U(Result i (A))

MEU: maximum expected utility A rational agent should choose an action that maximizes the agent’s expected utility This is a framework where all of the components of an AI system fit

16.2 The Basis of Utility Theory 6 Preference Axioms Orderability (A > B) or (B > A) or (A ~ B) Transitivity (A > B)  (B > C) -> (A > C) Continuity (A > B > C) ->  p [p, A; 1-p, C] ~ B (“lottery”) Substitutability (A ~ B) -> [p, A; 1-p, C] ~ [p, B; 1-p, C]

Monotonicity (A > B) -> (p >= q [p, A; 1-p, B] >~ [q, A; 1-q, B] Decomposability [p, A, 1-p, [q, B; 1-q, C]] ~ [p, A; (1-p)q, B; (1-p)(1-q), C]

Utility Principle If an agent’s preferences obey the axioms of utility, then there exists a real-valued function U that operates on states such that U(A) > U(B) if and only if A is preferred to B and U(A) = U(B) if and only if the agent is indifferent to A and B. U(A) > U(B) A > B U(A) = U(B) A ~ B

Maximum Expected Utility Principle The utility of a lottery is the sum of the probability of each outcome times the utility of that outcome. U([p 1, S 1 ; … ; p n, S n ]) = ∑ p i U(S i )

16.3 Utility Functions Utility of money, Figure 16.2 Monotonic preference Expected Monetary Value, EMV –take $1000 or 50% chance of $3000 –EU(Accept) =.5 *U(S k ) +.5*U(S k+3000 ) –EU(Decline) = U(S k+1000 )

Kahneman and Tversky (1982) A: 80% chance of $4000 B: 100% chance of $3000 Subjects prefer B,.8U(4000) < U(3000) C: 20% chance of $4000 D: 25% chance of $3000 Subjects prefer C,.2U(4000) >.25U(3000)

Utility Scales Utility functions are not unique U’(S) = k 1 + k 2 * U(S), k 2 > 0 u┬, best possible outcome, 1 u┴, worst possible outcome, 0 Micromort (1:1,000,000 chance of death), $20 in 1980 QALY: quality-adjusted life year