Uncertain Reasoning. Environment Description. Configurations. Models. Bayesian Networks

Size: px
Start display at page:

Download "Uncertain Reasoning. Environment Description. Configurations. Models. Bayesian Networks"

Transcription

1 Bayesian Networks A. Objectives 1. Basics on Bayesian probability theory 2. Belief updating using JPD 3. Basics on graphs 4. Bayesian networks 5. Acquisition of Bayesian networks 6. Local computation and message passing B. References a. Xiang (Chapter 2) b. Jensen & Nielsen (Chapter 3) c. Darwiche (2.3) d. Fenton & Neil (Chapter 7) Y. Xiang, CIS 6120, Uncertain Reasoning 1 Environment Description A. An agent can represent the state of its environment through a set V of variables. B. Ex A simple digital circuit C. Each variable x V is associated with a domain D x. D x may be discrete or continuous. Y. Xiang, CIS 6120, Uncertain Reasoning 2 Configurations A. We extend propositional logic to allow variables and AtomicSentence x = x i, where x i D x. B. Write a conjunctive sentence (a = a i ) (z = z k ) over a subset X V of variables as a configuration x = (a = a i,, z = z k ) or x = (a i,, z k ). 1. What does it represent? 2. It is also called a variable assignment or instantiation. C. The space of a subset X V of variables is the set D X of all possible configurations over X. D X = Cartesian product of domains of variables in X Models A. A model (or possible world) is a particular state of the environment. A model m is a function that maps each variable x to a value m(x) D x. B. Ex m 0 : a=hi, b=lo, c=hi, d= hi, e=lo, f=lo, h=hi, and g i =norm for i = 1, 2, 3, 4. C. A model is a complete (over V) configuration. D. Denote the set of all models of an env by. Ex For the circuit env, what is? Y. Xiang, CIS 6120, Uncertain Reasoning 3 Y. Xiang, CIS 6120, Uncertain Reasoning 4 1

2 Sentences and Models A. From a model m, the truth of any sentence is uniquely determined. Ex = (g 4 = norm) (c=hi) (d=lo). What is the truth of in m 0? B. If a sentence is true in a model m, we say that m satisfies or m is a model of. Ex Relation btw sentence and model m 0 C. Denote the set of all models of by M( ). 1. Ex What is M( )? 2. Illustration of M( ) by Venn diagram Y. Xiang, CIS 6120, Uncertain Reasoning 5 Logic and Set Operations on Models A. For sentences and, the following holds. 1. M( ) = M( ) M( ) 2. M( ) = M( ) M( ) 3. M( ) = \M( ) = M(α) 4. and are equivalent iff M( ) = M( ). 5. and are mutually exclusive iff M( ) M( ) =. 6. and are exhaustive iff M( ) M( ) =. 7. iff M( ) M( ). Y. Xiang, CIS 6120, Uncertain Reasoning 6 Possible and Compatible Configurations A. A model or configuration is impossible if it violates dependency among its variables. B. Configurations x and y are compatible iff for every v X Y, its value in x equals its value in y. 1) Can two models of an env be compatible? 2) If x and y are compatible, what is the relation between M(x) and M(y)? 3) What if x and y are incompatible? Encoding Uncertain Knowledge A. To infer the state of an environment, we rely on dependence btw observable and unobservable. How do we represent uncertain dependence? B. Bayesian approach: P(X=x Y=y) 1. Function P: D X D Y [0,1] 2. What does it mean? C. Ex Patient w walks into the office of doctor z. What is p = P z (w has TB B z )? D. Ex Patient w walks in with a positive X-ray film. What is q = P z (w has TB B z )? Y. Xiang, CIS 6120, Uncertain Reasoning 7 Y. Xiang, CIS 6120, Uncertain Reasoning 8 2

3 Probability Distributions 1. Conditional probability P(x y) 2. Conditional probability distribution (CPD) P(X y) 3. Conditional probability table (CPT) P(X Y) 4. Unconditional probability (distribution) P(X) 5. Joint probability distribution (JPD) P(V) 6. Marginal probability distribution P(X) Ex Conditional Probability Table idx h a e g 2 P(h a,e,g 2 ) idx h a e g 2 P(h a,e,g 2 ) norm abnm norm abnm norm abnm norm abnm norm abnm norm abnm norm abnm norm abnm 0.8 Y. Xiang, CIS 6120, Uncertain Reasoning 9 Y. Xiang, CIS 6120, Uncertain Reasoning 10 Axioms of Probability A. Probability assignment must satisfy a set of axioms. 1. X, Y, Z: nonempty, disjoint subsets of V 2. x D X, y D Y, and z D Z B. Range: 0 P(x y) 1 C. Certainty: P(x x) = 1 D. Sum: If x 1 and x 2 are incompatible, P(x 1 x 2 z) = P(x 1 z) + P(x 2 z). E. Product: P(x,y z) = P(x y,z) P(y z) Since it holds for all x, y and z, we often write P(X,Y Z) = P(X Y,Z) P(Y Z). Theorems of Probability A. From axioms, rest of probability theory is derivable. B. Bayes rule: P(X Y,Z) = P(Y X, Z)P(X Z)/P(Y Z). C. Negation: P(X x y) = 1 - P(X=x y). D. Marginalization: For Y = X W and X W =, P(x z) = P( x, w z). w D w Since the above holds for all x, w and z, we write P(X Z) = W P(X,W Z) = W P(Y Z). Y. Xiang, CIS 6120, Uncertain Reasoning 11 Y. Xiang, CIS 6120, Uncertain Reasoning 12 3

4 Probabilistic Inference A. Reasoning task: What is the state of agent env given the percept (observations)? B. The Bayesian theory distinguishes between the prior distribution over X V and the posterior distribution over X after observing Y = y. C. Probabilistic inference is the process to update prior into posterior after observation. D. In practice, probabilistic inference is performed recursively. Ex Dental Hygiene A. Bad habit in tooth cleaning increases chances of cavity which in turn cause toothache. B. V={hygiene, cavity, toothache}, where hygiene {g, b}, and cavity, toothache {y, n}. C. Given P(h, c, t) and percept toothache = yes, what is the posterior P(h t=y)? h c t P(h,c,t) good Yes yes good Yes no good No yes good no no bad yes yes bad yes no bad no yes bad no no Y. Xiang, CIS 6120, Uncertain Reasoning 13 Y. Xiang, CIS 6120, Uncertain Reasoning 14 Inference Using JPD A. Query: What is P(h t=y)? B. Algorithm 1. Constrain P(h,c,t) into P(h,c,t=y). 2. Compute P(t=y) by marginalization. 3. Condition P(h,c,t=y) into P(h,c,t t=y) by product. 4. Marginalize out c and t to get P(h t=y). C. Normalization: Steps 2 and 3 above Normalization constant: = 1/P(t=y) D. Hence, P(h t=y) = c,t P(h,c,t=y). Complexity of Inference Using JPD A. Suppose V =n, D x k for x V, and V =V\{x}. B. How many probability parameters are needed to specify P(V) in O() notation? Acquisition intractability C. How may parameters in P(V) define P(V x=x 0 )? 1. Agent must process these parameters to compute P(y x=x 0 ) for any y V. 2. Updating intractability D. How many summations are needed to get P(y x=x 0 ) from P(V x=x 0 )? Marginalization intractability Y. Xiang, Uncertain Reasoning with PGM 15 Y. Xiang, Uncertain Reasoning with PGM 16 4

5 Explore Independence A. Why is probabilistic inference using JPD intractable? B. Ex To know whether g 3 is normal, is it useful to observe c after observing b, d, and f? C. Observation: Not every env variable is directly dependent on everyone else. D. Ex If d is not observable, is it useful to observe c after observing only b and f? Encode Independence by Graph A. Key to efficient probabilistic inference Encode dependence (independence) relations among variables effectively B. Encoding dependence relations with graph 1) Direct dependence 2) Indirect dependence 3) Asymmetric or causal dependence 4) Unconditional independence 5) Conditional independence Y. Xiang, CIS 6120, Uncertain Reasoning 17 Y. Xiang, CIS 6120, Uncertain Reasoning 18 Basics on Graphs (Quick Review) 1) Graphs G = (V, E) 2) Undirected graphs: E = {<u,v> u,v V, u v} 3) Direct graphs: E = {(u,v) u,v V, u v} Head and tail; parent and child; 4) Degree, in-degree, and out-degree 5) Internal and terminal nodes; root and leaf; 6) Adjacent nodes and adjacency adj(v) 7) Path, simple path, directed path, ancestor and descendant 8) Cycles, directed, undirected, and DAGs 9) Singly connected and multiply connected graphs 10)Subgraphs Y. Xiang, CIS 6120, Uncertain Reasoning 19 Conditional Independence (CI) A. Let X, Y, and Z be disjoint subsets of variables in V. X and Y are conditionally independent given Z, denoted by I(X,Z,Y), iff for every x, and for every y and z such that P(y,z)>0, we have P(x y,z) = P(x z). B. Order of X & Y is unimportant: I(X,Z,Y) I(Y,Z,X). C. Only-if part is key to tackle acquisition intractability. D. Ex Dental hygiene: Acquire P(h,c,t) given I(h,c,t). E. If part yields a numerical test of CI, but cannot be relied on to tackle acquisition intractability. F. Ex Does I(h,c,t) holds for the given P(h,c,t)? Y. Xiang, CIS 6120, Uncertain Reasoning 20 5

6 Establishing Conditional Independence A. Without numerical test, how can we establish CI? B. Causality Given direct causes, a variable is conditionally independent of all others except its own effects. C. Mental test of CI D. Tractable acquisition 1) Identify CI by causality. 2) Acquire local distributions over cause & effect. 3) Define JPD from CI and local distributions. Encoding CI with DAG A. A Bayesian network is a triplet (V,G,P). 1. V is the set of variables of an environment. 2. G is a DAG whose nodes map one-to-one to members of V s.t. each variable is conditionally independent of its non-descendants given its parents. 3. P = { P(v (v)) v V} is a set of CPTs where (v) is the set of parents of v in G. B. Ex BN for a digital circuit Y. Xiang, CIS 6120, Uncertain Reasoning 21 Y. Xiang, CIS 6120, Uncertain Reasoning 22 An Example BN 1. How many parameters are needed to specify JPD directly? 2. How many are needed to specify BN? 3. How do we get JPD from the BN? Y. Xiang, CIS 6120, Uncertain Reasoning 23 JPD of a Bayesian Network A. Theorem [Chain rule] Let (V,G, P) be a BN, then P(V) = v V P(v (v)). B. Proof sketch (induction on V ) 1. Base case for V =1; Inductive assumption for V = n-1; 2. General case for V = n; Pick leaf a; Define BNs S and S ; 3. Apply inductive assumption to P(V ) 4. Apply product rule to P(V) C. A BN uniquely and compactly specifies a JPD. D. Is acquisition intractability resolved? Y. Xiang, CIS 6120, Uncertain Reasoning 24 6

7 How To Specify Bayesian Networks A. Main steps 1. Identify hypothesis variables 2. Define dependence structure 3. Determine variable domains 4. Specify CPTs B. Identify unobservable variables of primary concern a. Diagnosis: diseases of patient b. Web search: desirability of web page for given query c. Tutoring: knowledge state of student C. These variables drive structure specification. Defining Network Structure A. Specify a large network by divide and conquer. 1. Specify a number of network fragments. 2. Connect fragments into a single structure. B. Fragment specification can often be aided by reusing certain natural patterns (idioms). a. An idiom consists of a set of generic variables connected into a DAG. b. An idiom instantiation is an idiom with variables relabeled according to a specific application. Y. Xiang, CIS 6120, Uncertain Reasoning 25 Y. Xiang, CIS 6120, Uncertain Reasoning 26 Cause-Consequence Idiom A. Model a causal relation between a cause and its consequence. cause consequence B. Ex Idiom instantiations 1) Productive 2) Physical 3) Intentional C. An instantiation may have multiple causes. D. Support both predictive and diagnostic inference. Measurement Idiom A. Model uncertainty on an attribute due to sensor inaccuracy. attributevalue sensoraccuracy sensedvalue B. Ex Idiom instantiation on product testing C. Support diagnostic inference. Y. Xiang, CIS 6120, Uncertain Reasoning 27 Y. Xiang, CIS 6120, Uncertain Reasoning 28 7

8 Synthesis Idiom A. Model definitional relations. factor_1 factor_n syntheticfactor B. Ex Idiom instantiation on system reliability C. The relation may be deterministic or uncertain. D. Ex Instantiation on velocity Induction Idiom A. Model learning about a population parameter, in order to make predictions about it in the future, taking into account of contextual differences. populationparameter context obs_1 obs_n forecast B. Ex Idiom instantiation on call center monitoring A manager monitors lost calls for 8 days and predicts lost calls for the 9 th day. Y. Xiang, CIS 6120, Uncertain Reasoning 29 Y. Xiang, CIS 6120, Uncertain Reasoning 30 Specifying Variable Domains A. Associate each variable with a set of values that are mutually exclusive and exhaustive. 1) Testing mutual exclusiveness 2) Testing exhaustiveness B. Clarity criterion a) Values must be defined sufficiently precisely so that which is true can be practically determined. b) Ex Body Temperature Y. Xiang, CIS 6120, Uncertain Reasoning 31 Evaluating Probability Distributions 1. Follow well-founded theory if one exists. Ex One s genotype is aa (property shows), aa (carrier), or AA (non-carrier). A child takes one gene (a or A) from father s genotype (F) and one from mother s (M) with equal chance. What is P(C F,M)? F\M aa aa AA aa aa AA C aa aa AA aa aa AA aa aa AA 2. Use empirical frequency if available. 3. Use subjective estimate when necessary. Y. Xiang, CIS 6120, Uncertain Reasoning 32 8

9 Belief Updating By Message Passing A. Graphical structures help to copy with acquisition intractability. Can they help to copy with updating intractability and marginalization intractability? B. General idea 1. Avoid updating JPD directly and marginalizing it. 2. Update local distributions and propagate impact by passing messages along graphical structures. Ex Message Passing in Chain A. Ex BN for toothache P(h=good)=0.7 habit cavity p(c=yes h=good)=0.1 p(c=yes h=bad)=0.8 B. Compute P(h t=y) by message passing 1. Message P(t=y c) from t to c 2. Message P(t=y h) from c to h 3. Compute result at h p(t=yes c=yes)=0.85 p(t=yes c=no)=0.05 toothache Y. Xiang, CIS 6120, Uncertain Reasoning 33 Y. Xiang, CIS 6120, Uncertain Reasoning 34 Message Passing in Trees A. The - message passing [Pearl 88] computes the posterior for every node in a tree-structured BN with linear time. B. Not all envs admit tree dependence structures. Ex BN for a digital circuit C. When - message passing is applied to multiply connected BNs, there is no theoretical guarantee for exact posteriors. D. Experimental study obtained posteriors highly correlated with correct posteriors for some BNs but oscillated outcomes for others. Y. Xiang, CIS 6120, Uncertain Reasoning 35 Demo: Building Alarm A. Ex A building is typically normal but tampering or fire occurs occasionally. Either may trigger alarm which prompts occupants to evacuate. Verbal reports may reach security guard when gathering occurs outside the building. B. Separate inference engine from knowledge base C. Predictive inference D. Diagnostic inference E. Explaining away Y. Xiang, CIS 6120, Uncertain Reasoning 36 9

Y. Xiang, Inference with Uncertain Knowledge 1

Y. Xiang, Inference with Uncertain Knowledge 1 Inference with Uncertain Knowledge Objectives Why must agent use uncertain knowledge? Fundamentals of Bayesian probability Inference with full joint distributions Inference with Bayes rule Bayesian networks

More information

1. what conditional independencies are implied by the graph. 2. whether these independecies correspond to the probability distribution

1. what conditional independencies are implied by the graph. 2. whether these independecies correspond to the probability distribution NETWORK ANALYSIS Lourens Waldorp PROBABILITY AND GRAPHS The objective is to obtain a correspondence between the intuitive pictures (graphs) of variables of interest and the probability distributions of

More information

COS402- Artificial Intelligence Fall Lecture 10: Bayesian Networks & Exact Inference

COS402- Artificial Intelligence Fall Lecture 10: Bayesian Networks & Exact Inference COS402- Artificial Intelligence Fall 2015 Lecture 10: Bayesian Networks & Exact Inference Outline Logical inference and probabilistic inference Independence and conditional independence Bayes Nets Semantics

More information

Uncertainty and Bayesian Networks

Uncertainty and Bayesian Networks Uncertainty and Bayesian Networks Tutorial 3 Tutorial 3 1 Outline Uncertainty Probability Syntax and Semantics for Uncertainty Inference Independence and Bayes Rule Syntax and Semantics for Bayesian Networks

More information

Quantifying uncertainty & Bayesian networks

Quantifying uncertainty & Bayesian networks Quantifying uncertainty & Bayesian networks CE417: Introduction to Artificial Intelligence Sharif University of Technology Spring 2016 Soleymani Artificial Intelligence: A Modern Approach, 3 rd Edition,

More information

Probabilistic Reasoning. (Mostly using Bayesian Networks)

Probabilistic Reasoning. (Mostly using Bayesian Networks) Probabilistic Reasoning (Mostly using Bayesian Networks) Introduction: Why probabilistic reasoning? The world is not deterministic. (Usually because information is limited.) Ways of coping with uncertainty

More information

UNCERTAINTY. In which we see what an agent should do when not all is crystal-clear.

UNCERTAINTY. In which we see what an agent should do when not all is crystal-clear. UNCERTAINTY In which we see what an agent should do when not all is crystal-clear. Outline Uncertainty Probabilistic Theory Axioms of Probability Probabilistic Reasoning Independency Bayes Rule Summary

More information

Computer Science CPSC 322. Lecture 18 Marginalization, Conditioning

Computer Science CPSC 322. Lecture 18 Marginalization, Conditioning Computer Science CPSC 322 Lecture 18 Marginalization, Conditioning Lecture Overview Recap Lecture 17 Joint Probability Distribution, Marginalization Conditioning Inference by Enumeration Bayes Rule, Chain

More information

Introduction to Artificial Intelligence. Unit # 11

Introduction to Artificial Intelligence. Unit # 11 Introduction to Artificial Intelligence Unit # 11 1 Course Outline Overview of Artificial Intelligence State Space Representation Search Techniques Machine Learning Logic Probabilistic Reasoning/Bayesian

More information

COMP5211 Lecture Note on Reasoning under Uncertainty

COMP5211 Lecture Note on Reasoning under Uncertainty COMP5211 Lecture Note on Reasoning under Uncertainty Fangzhen Lin Department of Computer Science and Engineering Hong Kong University of Science and Technology Fangzhen Lin (HKUST) Uncertainty 1 / 33 Uncertainty

More information

Lecture 10: Introduction to reasoning under uncertainty. Uncertainty

Lecture 10: Introduction to reasoning under uncertainty. Uncertainty Lecture 10: Introduction to reasoning under uncertainty Introduction to reasoning under uncertainty Review of probability Axioms and inference Conditional probability Probability distributions COMP-424,

More information

Reasoning Under Uncertainty: Belief Network Inference

Reasoning Under Uncertainty: Belief Network Inference Reasoning Under Uncertainty: Belief Network Inference CPSC 322 Uncertainty 5 Textbook 10.4 Reasoning Under Uncertainty: Belief Network Inference CPSC 322 Uncertainty 5, Slide 1 Lecture Overview 1 Recap

More information

Tópicos Especiais em Modelagem e Análise - Aprendizado por Máquina CPS863

Tópicos Especiais em Modelagem e Análise - Aprendizado por Máquina CPS863 Tópicos Especiais em Modelagem e Análise - Aprendizado por Máquina CPS863 Daniel, Edmundo, Rosa Terceiro trimestre de 2012 UFRJ - COPPE Programa de Engenharia de Sistemas e Computação Bayesian Networks

More information

Probabilistic Robotics

Probabilistic Robotics Probabilistic Robotics Overview of probability, Representing uncertainty Propagation of uncertainty, Bayes Rule Application to Localization and Mapping Slides from Autonomous Robots (Siegwart and Nourbaksh),

More information

Intelligent Systems: Reasoning and Recognition. Reasoning with Bayesian Networks

Intelligent Systems: Reasoning and Recognition. Reasoning with Bayesian Networks Intelligent Systems: Reasoning and Recognition James L. Crowley ENSIMAG 2 / MoSIG M1 Second Semester 2016/2017 Lesson 13 24 march 2017 Reasoning with Bayesian Networks Naïve Bayesian Systems...2 Example

More information

Quantifying Uncertainty & Probabilistic Reasoning. Abdulla AlKhenji Khaled AlEmadi Mohammed AlAnsari

Quantifying Uncertainty & Probabilistic Reasoning. Abdulla AlKhenji Khaled AlEmadi Mohammed AlAnsari Quantifying Uncertainty & Probabilistic Reasoning Abdulla AlKhenji Khaled AlEmadi Mohammed AlAnsari Outline Previous Implementations What is Uncertainty? Acting Under Uncertainty Rational Decisions Basic

More information

Probabilistic Models

Probabilistic Models Bayes Nets 1 Probabilistic Models Models describe how (a portion of) the world works Models are always simplifications May not account for every variable May not account for all interactions between variables

More information

Recall from last time: Conditional probabilities. Lecture 2: Belief (Bayesian) networks. Bayes ball. Example (continued) Example: Inference problem

Recall from last time: Conditional probabilities. Lecture 2: Belief (Bayesian) networks. Bayes ball. Example (continued) Example: Inference problem Recall from last time: Conditional probabilities Our probabilistic models will compute and manipulate conditional probabilities. Given two random variables X, Y, we denote by Lecture 2: Belief (Bayesian)

More information

13.4 INDEPENDENCE. 494 Chapter 13. Quantifying Uncertainty

13.4 INDEPENDENCE. 494 Chapter 13. Quantifying Uncertainty 494 Chapter 13. Quantifying Uncertainty table. In a realistic problem we could easily have n>100, makingo(2 n ) impractical. The full joint distribution in tabular form is just not a practical tool for

More information

Directed and Undirected Graphical Models

Directed and Undirected Graphical Models Directed and Undirected Davide Bacciu Dipartimento di Informatica Università di Pisa bacciu@di.unipi.it Machine Learning: Neural Networks and Advanced Models (AA2) Last Lecture Refresher Lecture Plan Directed

More information

Bayesian Networks. Vibhav Gogate The University of Texas at Dallas

Bayesian Networks. Vibhav Gogate The University of Texas at Dallas Bayesian Networks Vibhav Gogate The University of Texas at Dallas Intro to AI (CS 4365) Many slides over the course adapted from either Dan Klein, Luke Zettlemoyer, Stuart Russell or Andrew Moore 1 Outline

More information

Bayesian networks. Soleymani. CE417: Introduction to Artificial Intelligence Sharif University of Technology Spring 2018

Bayesian networks. Soleymani. CE417: Introduction to Artificial Intelligence Sharif University of Technology Spring 2018 Bayesian networks CE417: Introduction to Artificial Intelligence Sharif University of Technology Spring 2018 Soleymani Slides have been adopted from Klein and Abdeel, CS188, UC Berkeley. Outline Probability

More information

Lecture 4 October 18th

Lecture 4 October 18th Directed and undirected graphical models Fall 2017 Lecture 4 October 18th Lecturer: Guillaume Obozinski Scribe: In this lecture, we will assume that all random variables are discrete, to keep notations

More information

Probabilistic Reasoning

Probabilistic Reasoning Course 16 :198 :520 : Introduction To Artificial Intelligence Lecture 7 Probabilistic Reasoning Abdeslam Boularias Monday, September 28, 2015 1 / 17 Outline We show how to reason and act under uncertainty.

More information

PROBABILISTIC REASONING SYSTEMS

PROBABILISTIC REASONING SYSTEMS PROBABILISTIC REASONING SYSTEMS In which we explain how to build reasoning systems that use network models to reason with uncertainty according to the laws of probability theory. Outline Knowledge in uncertain

More information

Bayesian Networks. Vibhav Gogate The University of Texas at Dallas

Bayesian Networks. Vibhav Gogate The University of Texas at Dallas Bayesian Networks Vibhav Gogate The University of Texas at Dallas Intro to AI (CS 6364) Many slides over the course adapted from either Dan Klein, Luke Zettlemoyer, Stuart Russell or Andrew Moore 1 Outline

More information

Bayesian Networks. Probability Theory and Probabilistic Reasoning. Emma Rollon and Javier Larrosa Q

Bayesian Networks. Probability Theory and Probabilistic Reasoning. Emma Rollon and Javier Larrosa Q Bayesian Networks Probability Theory and Probabilistic Reasoning Emma Rollon and Javier Larrosa Q1-2015-2016 Emma Rollon and Javier Larrosa Bayesian Networks Q1-2015-2016 1 / 26 Degrees of belief Given

More information

Modeling and reasoning with uncertainty

Modeling and reasoning with uncertainty CS 2710 Foundations of AI Lecture 18 Modeling and reasoning with uncertainty Milos Hauskrecht milos@cs.pitt.edu 5329 Sennott Square KB systems. Medical example. We want to build a KB system for the diagnosis

More information

Introduction to Probabilistic Graphical Models

Introduction to Probabilistic Graphical Models Introduction to Probabilistic Graphical Models Kyu-Baek Hwang and Byoung-Tak Zhang Biointelligence Lab School of Computer Science and Engineering Seoul National University Seoul 151-742 Korea E-mail: kbhwang@bi.snu.ac.kr

More information

Uncertainty. Introduction to Artificial Intelligence CS 151 Lecture 2 April 1, CS151, Spring 2004

Uncertainty. Introduction to Artificial Intelligence CS 151 Lecture 2 April 1, CS151, Spring 2004 Uncertainty Introduction to Artificial Intelligence CS 151 Lecture 2 April 1, 2004 Administration PA 1 will be handed out today. There will be a MATLAB tutorial tomorrow, Friday, April 2 in AP&M 4882 at

More information

Artificial Intelligence CS 6364

Artificial Intelligence CS 6364 Artificial Intelligence CS 6364 rofessor Dan Moldovan Section 12 robabilistic Reasoning Acting under uncertainty Logical agents assume propositions are - True - False - Unknown acting under uncertainty

More information

Introduction to Mobile Robotics Probabilistic Robotics

Introduction to Mobile Robotics Probabilistic Robotics Introduction to Mobile Robotics Probabilistic Robotics Wolfram Burgard 1 Probabilistic Robotics Key idea: Explicit representation of uncertainty (using the calculus of probability theory) Perception Action

More information

Course Introduction. Probabilistic Modelling and Reasoning. Relationships between courses. Dealing with Uncertainty. Chris Williams.

Course Introduction. Probabilistic Modelling and Reasoning. Relationships between courses. Dealing with Uncertainty. Chris Williams. Course Introduction Probabilistic Modelling and Reasoning Chris Williams School of Informatics, University of Edinburgh September 2008 Welcome Administration Handout Books Assignments Tutorials Course

More information

Rapid Introduction to Machine Learning/ Deep Learning

Rapid Introduction to Machine Learning/ Deep Learning Rapid Introduction to Machine Learning/ Deep Learning Hyeong In Choi Seoul National University 1/32 Lecture 5a Bayesian network April 14, 2016 2/32 Table of contents 1 1. Objectives of Lecture 5a 2 2.Bayesian

More information

CSE 473: Artificial Intelligence Autumn 2011

CSE 473: Artificial Intelligence Autumn 2011 CSE 473: Artificial Intelligence Autumn 2011 Bayesian Networks Luke Zettlemoyer Many slides over the course adapted from either Dan Klein, Stuart Russell or Andrew Moore 1 Outline Probabilistic models

More information

Directed Graphical Models

Directed Graphical Models CS 2750: Machine Learning Directed Graphical Models Prof. Adriana Kovashka University of Pittsburgh March 28, 2017 Graphical Models If no assumption of independence is made, must estimate an exponential

More information

Bayesian Networks: Representation, Variable Elimination

Bayesian Networks: Representation, Variable Elimination Bayesian Networks: Representation, Variable Elimination CS 6375: Machine Learning Class Notes Instructor: Vibhav Gogate The University of Texas at Dallas We can view a Bayesian network as a compact representation

More information

An AI-ish view of Probability, Conditional Probability & Bayes Theorem

An AI-ish view of Probability, Conditional Probability & Bayes Theorem An AI-ish view of Probability, Conditional Probability & Bayes Theorem Review: Uncertainty and Truth Values: a mismatch Let action A t = leave for airport t minutes before flight. Will A 15 get me there

More information

10/18/2017. An AI-ish view of Probability, Conditional Probability & Bayes Theorem. Making decisions under uncertainty.

10/18/2017. An AI-ish view of Probability, Conditional Probability & Bayes Theorem. Making decisions under uncertainty. An AI-ish view of Probability, Conditional Probability & Bayes Theorem Review: Uncertainty and Truth Values: a mismatch Let action A t = leave for airport t minutes before flight. Will A 15 get me there

More information

Probabilistic Reasoning. Kee-Eung Kim KAIST Computer Science

Probabilistic Reasoning. Kee-Eung Kim KAIST Computer Science Probabilistic Reasoning Kee-Eung Kim KAIST Computer Science Outline #1 Acting under uncertainty Probabilities Inference with Probabilities Independence and Bayes Rule Bayesian networks Inference in Bayesian

More information

Machine Learning Lecture 14

Machine Learning Lecture 14 Many slides adapted from B. Schiele, S. Roth, Z. Gharahmani Machine Learning Lecture 14 Undirected Graphical Models & Inference 23.06.2015 Bastian Leibe RWTH Aachen http://www.vision.rwth-aachen.de/ leibe@vision.rwth-aachen.de

More information

Bayesian Networks and Decision Graphs

Bayesian Networks and Decision Graphs Bayesian Networks and Decision Graphs A 3-week course at Reykjavik University Finn V. Jensen & Uffe Kjærulff ({fvj,uk}@cs.aau.dk) Group of Machine Intelligence Department of Computer Science, Aalborg University

More information

Directed Graphical Models or Bayesian Networks

Directed Graphical Models or Bayesian Networks Directed Graphical Models or Bayesian Networks Le Song Machine Learning II: Advanced Topics CSE 8803ML, Spring 2012 Bayesian Networks One of the most exciting recent advancements in statistical AI Compact

More information

Motivation. Bayesian Networks in Epistemology and Philosophy of Science Lecture. Overview. Organizational Issues

Motivation. Bayesian Networks in Epistemology and Philosophy of Science Lecture. Overview. Organizational Issues Bayesian Networks in Epistemology and Philosophy of Science Lecture 1: Bayesian Networks Center for Logic and Philosophy of Science Tilburg University, The Netherlands Formal Epistemology Course Northern

More information

Uncertainty. Logic and Uncertainty. Russell & Norvig. Readings: Chapter 13. One problem with logical-agent approaches: C:145 Artificial

Uncertainty. Logic and Uncertainty. Russell & Norvig. Readings: Chapter 13. One problem with logical-agent approaches: C:145 Artificial C:145 Artificial Intelligence@ Uncertainty Readings: Chapter 13 Russell & Norvig. Artificial Intelligence p.1/43 Logic and Uncertainty One problem with logical-agent approaches: Agents almost never have

More information

Fusion in simple models

Fusion in simple models Fusion in simple models Peter Antal antal@mit.bme.hu A.I. February 8, 2018 1 Basic concepts of probability theory Joint distribution Conditional probability Bayes rule Chain rule Marginalization General

More information

Informatics 2D Reasoning and Agents Semester 2,

Informatics 2D Reasoning and Agents Semester 2, Informatics 2D Reasoning and Agents Semester 2, 2017 2018 Alex Lascarides alex@inf.ed.ac.uk Lecture 23 Probabilistic Reasoning with Bayesian Networks 15th March 2018 Informatics UoE Informatics 2D 1 Where

More information

COMP9414: Artificial Intelligence Reasoning Under Uncertainty

COMP9414: Artificial Intelligence Reasoning Under Uncertainty COMP9414, Monday 16 April, 2012 Reasoning Under Uncertainty 2 COMP9414: Artificial Intelligence Reasoning Under Uncertainty Overview Problems with Logical Approach What Do the Numbers Mean? Wayne Wobcke

More information

Probabilistic Models. Models describe how (a portion of) the world works

Probabilistic Models. Models describe how (a portion of) the world works Probabilistic Models Models describe how (a portion of) the world works Models are always simplifications May not account for every variable May not account for all interactions between variables All models

More information

Outline. CSE 573: Artificial Intelligence Autumn Bayes Nets: Big Picture. Bayes Net Semantics. Hidden Markov Models. Example Bayes Net: Car

Outline. CSE 573: Artificial Intelligence Autumn Bayes Nets: Big Picture. Bayes Net Semantics. Hidden Markov Models. Example Bayes Net: Car CSE 573: Artificial Intelligence Autumn 2012 Bayesian Networks Dan Weld Many slides adapted from Dan Klein, Stuart Russell, Andrew Moore & Luke Zettlemoyer Outline Probabilistic models (and inference)

More information

Reasoning Under Uncertainty

Reasoning Under Uncertainty Reasoning Under Uncertainty Introduction Representing uncertain knowledge: logic and probability (a reminder!) Probabilistic inference using the joint probability distribution Bayesian networks The Importance

More information

Reasoning Under Uncertainty: Conditioning, Bayes Rule & the Chain Rule

Reasoning Under Uncertainty: Conditioning, Bayes Rule & the Chain Rule Reasoning Under Uncertainty: Conditioning, Bayes Rule & the Chain Rule Alan Mackworth UBC CS 322 Uncertainty 2 March 13, 2013 Textbook 6.1.3 Lecture Overview Recap: Probability & Possible World Semantics

More information

Probabilistic Graphical Models (I)

Probabilistic Graphical Models (I) Probabilistic Graphical Models (I) Hongxin Zhang zhx@cad.zju.edu.cn State Key Lab of CAD&CG, ZJU 2015-03-31 Probabilistic Graphical Models Modeling many real-world problems => a large number of random

More information

A Tutorial on Bayesian Belief Networks

A Tutorial on Bayesian Belief Networks A Tutorial on Bayesian Belief Networks Mark L Krieg Surveillance Systems Division Electronics and Surveillance Research Laboratory DSTO TN 0403 ABSTRACT This tutorial provides an overview of Bayesian belief

More information

Part I Qualitative Probabilistic Networks

Part I Qualitative Probabilistic Networks Part I Qualitative Probabilistic Networks In which we study enhancements of the framework of qualitative probabilistic networks. Qualitative probabilistic networks allow for studying the reasoning behaviour

More information

CS 188: Artificial Intelligence Fall 2008

CS 188: Artificial Intelligence Fall 2008 CS 188: Artificial Intelligence Fall 2008 Lecture 14: Bayes Nets 10/14/2008 Dan Klein UC Berkeley 1 1 Announcements Midterm 10/21! One page note sheet Review sessions Friday and Sunday (similar) OHs on

More information

Probabilistic representation and reasoning

Probabilistic representation and reasoning Probabilistic representation and reasoning Applied artificial intelligence (EDA132) Lecture 09 2017-02-15 Elin A. Topp Material based on course book, chapter 13, 14.1-3 1 Show time! Two boxes of chocolates,

More information

CS 5522: Artificial Intelligence II

CS 5522: Artificial Intelligence II CS 5522: Artificial Intelligence II Bayes Nets Instructor: Alan Ritter Ohio State University [These slides were adapted from CS188 Intro to AI at UC Berkeley. All materials available at http://ai.berkeley.edu.]

More information

Introduction to Bayesian Learning

Introduction to Bayesian Learning Course Information Introduction Introduction to Bayesian Learning Davide Bacciu Dipartimento di Informatica Università di Pisa bacciu@di.unipi.it Apprendimento Automatico: Fondamenti - A.A. 2016/2017 Outline

More information

Artificial Intelligence Uncertainty

Artificial Intelligence Uncertainty Artificial Intelligence Uncertainty Ch. 13 Uncertainty Let action A t = leave for airport t minutes before flight Will A t get me there on time? A 25, A 60, A 3600 Uncertainty: partial observability (road

More information

4 : Exact Inference: Variable Elimination

4 : Exact Inference: Variable Elimination 10-708: Probabilistic Graphical Models 10-708, Spring 2014 4 : Exact Inference: Variable Elimination Lecturer: Eric P. ing Scribes: Soumya Batra, Pradeep Dasigi, Manzil Zaheer 1 Probabilistic Inference

More information

Objectives. Probabilistic Reasoning Systems. Outline. Independence. Conditional independence. Conditional independence II.

Objectives. Probabilistic Reasoning Systems. Outline. Independence. Conditional independence. Conditional independence II. Copyright Richard J. Povinelli rev 1.0, 10/1//2001 Page 1 Probabilistic Reasoning Systems Dr. Richard J. Povinelli Objectives You should be able to apply belief networks to model a problem with uncertainty.

More information

CS 188: Artificial Intelligence Fall 2009

CS 188: Artificial Intelligence Fall 2009 CS 188: Artificial Intelligence Fall 2009 Lecture 14: Bayes Nets 10/13/2009 Dan Klein UC Berkeley Announcements Assignments P3 due yesterday W2 due Thursday W1 returned in front (after lecture) Midterm

More information

Introduction to Probabilistic Graphical Models

Introduction to Probabilistic Graphical Models Introduction to Probabilistic Graphical Models Franz Pernkopf, Robert Peharz, Sebastian Tschiatschek Graz University of Technology, Laboratory of Signal Processing and Speech Communication Inffeldgasse

More information

Bayesian Networks: Independencies and Inference

Bayesian Networks: Independencies and Inference Bayesian Networks: Independencies and Inference Scott Davies and Andrew Moore Note to other teachers and users of these slides. Andrew and Scott would be delighted if you found this source material useful

More information

CS6220: DATA MINING TECHNIQUES

CS6220: DATA MINING TECHNIQUES CS6220: DATA MINING TECHNIQUES Chapter 8&9: Classification: Part 3 Instructor: Yizhou Sun yzsun@ccs.neu.edu March 12, 2013 Midterm Report Grade Distribution 90-100 10 80-89 16 70-79 8 60-69 4

More information

Probabilistic Reasoning Systems

Probabilistic Reasoning Systems Probabilistic Reasoning Systems Dr. Richard J. Povinelli Copyright Richard J. Povinelli rev 1.0, 10/7/2001 Page 1 Objectives You should be able to apply belief networks to model a problem with uncertainty.

More information

Probabilistic Graphical Networks: Definitions and Basic Results

Probabilistic Graphical Networks: Definitions and Basic Results This document gives a cursory overview of Probabilistic Graphical Networks. The material has been gleaned from different sources. I make no claim to original authorship of this material. Bayesian Graphical

More information

Events A and B are independent P(A) = P(A B) = P(A B) / P(B)

Events A and B are independent P(A) = P(A B) = P(A B) / P(B) Events A and B are independent A B U P(A) = P(A B) = P(A B) / P(B) 1 Alternative Characterization of Independence a) P(A B) = P(A) b) P(A B) = P(A) P(B) Recall P(A B) = P(A B) / P(B) (if P(B) 0) So P(A

More information

Artificial Intelligence Programming Probability

Artificial Intelligence Programming Probability Artificial Intelligence Programming Probability Chris Brooks Department of Computer Science University of San Francisco Department of Computer Science University of San Francisco p.1/?? 13-0: Uncertainty

More information

A Decision Theoretic View on Choosing Heuristics for Discovery of Graphical Models

A Decision Theoretic View on Choosing Heuristics for Discovery of Graphical Models A Decision Theoretic View on Choosing Heuristics for Discovery of Graphical Models Y. Xiang University of Guelph, Canada Abstract Discovery of graphical models is NP-hard in general, which justifies using

More information

Ch.6 Uncertain Knowledge. Logic and Uncertainty. Representation. One problem with logical approaches: Department of Computer Science

Ch.6 Uncertain Knowledge. Logic and Uncertainty. Representation. One problem with logical approaches: Department of Computer Science Ch.6 Uncertain Knowledge Representation Hantao Zhang http://www.cs.uiowa.edu/ hzhang/c145 The University of Iowa Department of Computer Science Artificial Intelligence p.1/39 Logic and Uncertainty One

More information

6.867 Machine learning, lecture 23 (Jaakkola)

6.867 Machine learning, lecture 23 (Jaakkola) Lecture topics: Markov Random Fields Probabilistic inference Markov Random Fields We will briefly go over undirected graphical models or Markov Random Fields (MRFs) as they will be needed in the context

More information

Bayesian Network. Outline. Bayesian Network. Syntax Semantics Exact inference by enumeration Exact inference by variable elimination

Bayesian Network. Outline. Bayesian Network. Syntax Semantics Exact inference by enumeration Exact inference by variable elimination Outline Syntax Semantics Exact inference by enumeration Exact inference by variable elimination s A simple, graphical notation for conditional independence assertions and hence for compact specication

More information

CS 2750: Machine Learning. Bayesian Networks. Prof. Adriana Kovashka University of Pittsburgh March 14, 2016

CS 2750: Machine Learning. Bayesian Networks. Prof. Adriana Kovashka University of Pittsburgh March 14, 2016 CS 2750: Machine Learning Bayesian Networks Prof. Adriana Kovashka University of Pittsburgh March 14, 2016 Plan for today and next week Today and next time: Bayesian networks (Bishop Sec. 8.1) Conditional

More information

Statistical Approaches to Learning and Discovery

Statistical Approaches to Learning and Discovery Statistical Approaches to Learning and Discovery Graphical Models Zoubin Ghahramani & Teddy Seidenfeld zoubin@cs.cmu.edu & teddy@stat.cmu.edu CALD / CS / Statistics / Philosophy Carnegie Mellon University

More information

Reasoning Under Uncertainty

Reasoning Under Uncertainty Reasoning Under Uncertainty Introduction Representing uncertain knowledge: logic and probability (a reminder!) Probabilistic inference using the joint probability distribution Bayesian networks (theory

More information

Graphical Models. Andrea Passerini Statistical relational learning. Graphical Models

Graphical Models. Andrea Passerini Statistical relational learning. Graphical Models Andrea Passerini passerini@disi.unitn.it Statistical relational learning Probability distributions Bernoulli distribution Two possible values (outcomes): 1 (success), 0 (failure). Parameters: p probability

More information

CS 343: Artificial Intelligence

CS 343: Artificial Intelligence CS 343: Artificial Intelligence Bayes Nets Prof. Scott Niekum The University of Texas at Austin [These slides based on those of Dan Klein and Pieter Abbeel for CS188 Intro to AI at UC Berkeley. All CS188

More information

CMPSCI 240: Reasoning about Uncertainty

CMPSCI 240: Reasoning about Uncertainty CMPSCI 240: Reasoning about Uncertainty Lecture 17: Representing Joint PMFs and Bayesian Networks Andrew McGregor University of Massachusetts Last Compiled: April 7, 2017 Warm Up: Joint distributions Recall

More information

Belief Update in CLG Bayesian Networks With Lazy Propagation

Belief Update in CLG Bayesian Networks With Lazy Propagation Belief Update in CLG Bayesian Networks With Lazy Propagation Anders L Madsen HUGIN Expert A/S Gasværksvej 5 9000 Aalborg, Denmark Anders.L.Madsen@hugin.com Abstract In recent years Bayesian networks (BNs)

More information

Web-Mining Agents Data Mining

Web-Mining Agents Data Mining Web-Mining Agents Data Mining Prof. Dr. Ralf Möller Dr. Özgür L. Özçep Universität zu Lübeck Institut für Informationssysteme Tanya Braun (Übungen) 2 Uncertainty AIMA Chapter 13 3 Outline Agents Uncertainty

More information

Uncertainty. Chapter 13

Uncertainty. Chapter 13 Uncertainty Chapter 13 Outline Uncertainty Probability Syntax and Semantics Inference Independence and Bayes Rule Uncertainty Let s say you want to get to the airport in time for a flight. Let action A

More information

CS6220: DATA MINING TECHNIQUES

CS6220: DATA MINING TECHNIQUES CS6220: DATA MINING TECHNIQUES Matrix Data: Classification: Part 2 Instructor: Yizhou Sun yzsun@ccs.neu.edu September 21, 2014 Methods to Learn Matrix Data Set Data Sequence Data Time Series Graph & Network

More information

Probabilistic Graphical Models. Rudolf Kruse, Alexander Dockhorn Bayesian Networks 153

Probabilistic Graphical Models. Rudolf Kruse, Alexander Dockhorn Bayesian Networks 153 Probabilistic Graphical Models Rudolf Kruse, Alexander Dockhorn Bayesian Networks 153 The Big Objective(s) In a wide variety of application fields two main problems need to be addressed over and over:

More information

Reasoning Under Uncertainty: Bayesian networks intro

Reasoning Under Uncertainty: Bayesian networks intro Reasoning Under Uncertainty: Bayesian networks intro Alan Mackworth UBC CS 322 Uncertainty 4 March 18, 2013 Textbook 6.3, 6.3.1, 6.5, 6.5.1, 6.5.2 Lecture Overview Recap: marginal and conditional independence

More information

Foundations of Artificial Intelligence

Foundations of Artificial Intelligence Foundations of Artificial Intelligence 12. Making Simple Decisions under Uncertainty Probability Theory, Bayesian Networks, Other Approaches Wolfram Burgard, Maren Bennewitz, and Marco Ragni Albert-Ludwigs-Universität

More information

Preliminaries Bayesian Networks Graphoid Axioms d-separation Wrap-up. Bayesian Networks. Brandon Malone

Preliminaries Bayesian Networks Graphoid Axioms d-separation Wrap-up. Bayesian Networks. Brandon Malone Preliminaries Graphoid Axioms d-separation Wrap-up Much of this material is adapted from Chapter 4 of Darwiche s book January 23, 2014 Preliminaries Graphoid Axioms d-separation Wrap-up 1 Preliminaries

More information

Review: Bayesian learning and inference

Review: Bayesian learning and inference Review: Bayesian learning and inference Suppose the agent has to make decisions about the value of an unobserved query variable X based on the values of an observed evidence variable E Inference problem:

More information

Bayesian networks (1) Lirong Xia

Bayesian networks (1) Lirong Xia Bayesian networks (1) Lirong Xia Random variables and joint distributions Ø A random variable is a variable with a domain Random variables: capital letters, e.g. W, D, L values: small letters, e.g. w,

More information

Reasoning under Uncertainty: Intro to Probability

Reasoning under Uncertainty: Intro to Probability Reasoning under Uncertainty: Intro to Probability Computer Science cpsc322, Lecture 24 (Textbook Chpt 6.1, 6.1.1) March, 15, 2010 CPSC 322, Lecture 24 Slide 1 To complete your Learning about Logics Review

More information

CS 188: Artificial Intelligence Spring Announcements

CS 188: Artificial Intelligence Spring Announcements CS 188: Artificial Intelligence Spring 2011 Lecture 14: Bayes Nets II Independence 3/9/2011 Pieter Abbeel UC Berkeley Many slides over the course adapted from Dan Klein, Stuart Russell, Andrew Moore Announcements

More information

Artificial Intelligence

Artificial Intelligence Artificial Intelligence Dr Ahmed Rafat Abas Computer Science Dept, Faculty of Computers and Informatics, Zagazig University arabas@zu.edu.eg http://www.arsaliem.faculty.zu.edu.eg/ Uncertainty Chapter 13

More information

Bayes Nets: Independence

Bayes Nets: Independence Bayes Nets: Independence [These slides were created by Dan Klein and Pieter Abbeel for CS188 Intro to AI at UC Berkeley. All CS188 materials are available at http://ai.berkeley.edu.] Bayes Nets A Bayes

More information

Bayesian Networks BY: MOHAMAD ALSABBAGH

Bayesian Networks BY: MOHAMAD ALSABBAGH Bayesian Networks BY: MOHAMAD ALSABBAGH Outlines Introduction Bayes Rule Bayesian Networks (BN) Representation Size of a Bayesian Network Inference via BN BN Learning Dynamic BN Introduction Conditional

More information

Basic Probabilistic Reasoning SEG

Basic Probabilistic Reasoning SEG Basic Probabilistic Reasoning SEG 7450 1 Introduction Reasoning under uncertainty using probability theory Dealing with uncertainty is one of the main advantages of an expert system over a simple decision

More information

Bayesian networks. Chapter Chapter

Bayesian networks. Chapter Chapter Bayesian networks Chapter 14.1 3 Chapter 14.1 3 1 Outline Syntax Semantics Parameterized distributions Chapter 14.1 3 2 Bayesian networks A simple, graphical notation for conditional independence assertions

More information

Pengju

Pengju Introduction to AI Chapter13 Uncertainty Pengju Ren@IAIR Outline Uncertainty Probability Syntax and Semantics Inference Independence and Bayes Rule Example: Car diagnosis Wumpus World Environment Squares

More information

Reasoning with Bayesian Networks

Reasoning with Bayesian Networks Reasoning with Lecture 1: Probability Calculus, NICTA and ANU Reasoning with Overview of the Course Probability calculus, Bayesian networks Inference by variable elimination, factor elimination, conditioning

More information

Our Status. We re done with Part I Search and Planning!

Our Status. We re done with Part I Search and Planning! Probability [These slides were created by Dan Klein and Pieter Abbeel for CS188 Intro to AI at UC Berkeley. All CS188 materials are available at http://ai.berkeley.edu.] Our Status We re done with Part

More information