VIEWS: 8 PAGES: 66 POSTED ON: 3/2/2012 Public Domain
Abduction, Uncertainty, and Probabilistic Reasoning Chapters 13, 14, and more 1 Introduction • Abduction is a reasoning process that tries to form plausible explanations for abnormal observations – Abduction is distinct different from deduction and induction – Abduction is inherently uncertain • Uncertainty becomes an important issue in AI research • Some major formalisms for representing and reasoning about uncertainty – Mycin’s certainty factor (an early representative) – Probability theory (esp. Bayesian networks) – Dempster-Shafer theory – Fuzzy logic – Truth maintenance systems 2 Abduction • Definition (Encyclopedia Britannica): reasoning that derives an explanatory hypothesis from a given set of facts – The inference result is a hypothesis, which if true, could explain the occurrence of the given facts • Examples – Dendral, an expert system to construct 3D structure of chemical compounds • Fact: mass spectrometer data of the compound and the chemical formula of the compound • KB: chemistry, esp. strength of different types of bounds • Reasoning: form a hypothetical 3D structure which meets the given chemical formula, and would most likely produce the given mass spectrum if subjected to electron beam bombardment 3 – Medical diagnosis • Facts: symptoms, lab test results, and other observed findings (called manifestations) • KB: causal associations between diseases and manifestations • Reasoning: one or more diseases whose presence would causally explain the occurrence of the given manifestations – Many other reasoning processes (e.g., word sense disambiguation in natural language process, image understanding, detective’s work, etc.) can also been seen as abductive reasoning. 4 Comparing abduction, deduction and induction Deduction: major premise: All balls in the box are black A => B A minor premise: This ball is from the box --------- conclusion: This ball is black B Abduction: rule: All balls in the box are black A => B observation: This ball is black B ------------- explanation: This ball is from the box Possibly A Induction: case: These balls are from the box Whenever observation: These balls are black A then B but not hypothesized rule: All ball in the box are black vice versa ------------- Possibly A => B Induction: from specific cases to general rules Abduction and deduction: both from part of a specific case to other part of the case using general rules (in different ways) 5 Characteristics of abduction reasoning 1. Reasoning results are hypotheses, not theorems (may be false even if rules and facts are true), – e.g., misdiagnosis in medicine 2. There may be multiple plausible hypotheses – When given rules A => B and C => B, and fact B both A and C are plausible hypotheses – Abduction is inherently uncertain – Hypotheses can be ranked by their plausibility if that can be determined 3. Reasoning is often a Hypothesize-and-test cycle – hypothesize phase: postulate possible hypotheses, each of which could explain the given facts (or explain most of the important facts) – test phase: test the plausibility of all or some of these hypotheses 6 – One way to test a hypothesis H is to test if something that is currently unknown but can be predicted from H is actually true. • If we also know A => D and C => E, then ask if D and E are true. • If it turns out D is true and E is false, then hypothesis A becomes more plausible (support for A increased, support for C decreased) • Alternative hypotheses compete with each other (Okam’s razor prefers simpler hypotheses) 4. Reasoning is non-monotonic – Plausibility of hypotheses can increase/decrease as new facts are collected (deductive inference determines if a sentence is true but would never change its truth value) – Some hypotheses may be discarded/defeated, and new ones may be formed when new observations are made 7 Source of Uncertainty in Intelligent Systems • Uncertain data (noise) • Uncertain knowledge (e.g, causal relations) – A disorder may cause any and all POSSIBLE manifestations in a specific case – A manifestation can be caused by more than one POSSIBLE disorders • Uncertain reasoning results – Abduction and induction are inherently uncertain – Default reasoning, even in deductive fashion, is uncertain – Incomplete deductive inference may be uncertain 8 Probabilistic Inference • Based on probability theory (especially Bayes’ theorem) – Well established discipline about uncertain outcomes – Empirical science like physics/chemistry, can be verified by experiments • Probability theory is too rigid to apply directly in many knowledge-based applications – Some assumptions have to be made to simplify the reality – Different formalisms have been developed in which some aspects of the probability theory are changed/modified. • We will briefly review the basics of probability theory before discussing different approaches to uncertainty • The presentation uses diagnostic process (an abductive and evidential reasoning process) as an example 9 Probability of Events • Sample space and events – Sample space S: (e.g., all people in an area) – Events E1 S: (e.g., all people having cough) E2 S: (e.g., all people having cold) • Prior (marginal) probabilities of events – P(E) = |E| / |S| (frequency interpretation) – P(E) = 0.1 (subjective probability) – 0 <= P(E) <= 1 for all events – Two special events: and S: P() = 0 and P(S) = 1.0 • Boolean operators between events (to form compound events) – Conjunctive (intersection): E1 ^ E2 ( E1 E2) – Disjunctive (union): E1 v E2 ( E1 E2) – Negation (complement): ~E (E C = S – E) 10 • Probabilities of compound events – P(~E) = 1 – P(E) because P(~E) + P(E) =1 – P(E1 v E2) = P(E1) + P(E2) – P(E1 ^ E2) – But how to compute the joint probability P(E1 ^ E2)? ~E E E1 E2 E1 ^ E2 • Conditional probability (of E1, given E2) – How likely E1 occurs in the subspace of E2 | E1 E 2 | | E1 E 2 | / | S | P ( E1 E 2) P ( E1 | E 2) | E2 | | E2 | / | S | P ( E 2) P ( E1 E 2) P ( E1 | E 2) P ( E 2) 11 • Independence assumption – Two events E1 and E2 are said to be independent of each other if P ( E1 | E 2) P ( E1) (given E2 does not change the likelihood of E1) – Computation can be simplified with independent events P ( E1 E 2) P ( E1 | E 2) P ( E 2) P ( E1) P ( E 2) P ( E1 E 2) P ( E1) P ( E 2) P ( E1 E 2) P ( E1) P ( E 2) P ( E1) P ( E 2) 1 (1 P ( E1)(1 P ( E 2)) • Mutually exclusive (ME) and exhaustive (EXH) set of events – ME: E i E j ( P ( E i E j ) 0), i , j 1,.., n, i j – EXH: E1 ... E n S ( P ( E1 ... E n ) 1) 12 Bayes’ Theorem • In the setting of diagnostic/evidential reasoning H i P(H i ) hypotheses P(E j | Hi ) E1 Ej Em evidence/manifestations – Know prior probability of hypothesis P(H i ) conditional probability P(E j | Hi ) – Want to compute the posterior probability P ( H i | E j ) • Bayes’ theorem (formula 1): P ( H i | E j ) P ( H i ) P ( E j | H i ) / P ( E j ) • If the purpose is to find which of the n hypotheses H1 ,..., H n is more plausible given E j, then we can ignore the denominator and rank them use relative likelihood rel ( H i | E j ) P ( E j | H i ) P ( H i ) 13 • P ( E j ) can be computed from P ( E j | H i ) and P ( H i ) , if we assume all hypotheses H1 ,..., H n are ME and EXH P ( E j ) P ( E j ( H 1 ... H n ) ) (by EXH) n P( E j H i ) (by ME) i 1 n P( E j | H i )P( H i ) i 1 • Then we have another version of Bayes’ theorem: P(E j | Hi )P(Hi ) rel ( H i | E j ) P(Hi | E j ) n n P(E k 1 j | Hk )P(Hk ) rel ( H k 1 k | Ej) n where P(E k 1 j | H k ) P ( H k ) , the sum of relative likelihood of all n hypotheses, is a normalization factor 14 Probabilistic Inference for simple diagnostic problems • Knowledge base: E1 ,..., E m : evidence/manifestation H1 ,..., H n : hypotheses/disorders E j and H i are binary and hypotheses form a ME & EXH set P ( H i ), i 1,...n prior probabilit ies P ( E j | H i ), i 1,...n, j 1,...m conditiona l probabilit ies • Case input: E1 ,..., E l • Find the hypothesisH i with the highest posterior probability P ( H i | E1 ,..., E l ) P ( E1 ,... E l | H i ) P ( H i ) • By Bayes’ theorem P ( H i | E1 ,..., E l ) P ( E1 ,... E l ) • Assume all pieces of evidence are conditionally independent, given any hypothesis P( E1,...El | Hi ) lj 1P( E j | Hi ) 15 • The relative likelihood rel ( H i | E1 ,...,El ) P ( E1 ,...,El | H i ) P ( H i ) P ( H i )lj 1 P ( E j | H i ) • The absolute posterior probability rel ( H i | E1 ,...,E l ) P ( H i ) lj 1 P ( E j | H i ) P ( H i | E1 ,...,El ) P ( H k ) lj 1 P ( E j | H k ) n n rel ( H k | E1 ,...,El ) k 1 k 1 • Evidence accumulation (when new evidence discovered) If El+1 present rel ( H i | E1 ,..., E l , E l 1 ) P ( E l 1 | H i )rel ( H i | E1 ,..., E l ) If El+1 present rel ( H i | E1 ,..., E l , ~ E l 1 ) (1 P ( E l 1 | H i )) rel ( H i | E1 ,..., E l ) 16 Assessing the Assumptions • Assumption 1: hypotheses are mutually exclusive and exhaustive – Single fault assumption (one and only hypothesis must true) – Multi-faults do exist in individual cases – Can be viewed as an approximation of situations where hypotheses are independent of each other and their prior probabilities are very small P ( H1 H 2 ) P ( H1 ) P ( H 2 ) 0 if both P ( H1 ) and P ( H 2 ) are very small • Assumption 2: pieces of evidence are conditionally independent of each other, given any hypothesis – Manifestations themselves are not independent of each other, they are correlated by their common causes – Reasonable under single fault assumption – Not so when multi-faults are to be considered 17 Limitations of the simple Bayesian system • Cannot handle well hypotheses of multiple disorders – Suppose H1 ,..., H n are independent of each other – Consider a composite hypothesis H1 ^ H 2 – How to compute the posterior probability (or relative likelihood) P ( H1 ^ H 2 | E1 ,..., E l ) ? – Using Bayes’ theorem P ( E1 ,... E l | H1 ^ H 2 ) P ( H1 ^ H 2 ) P ( H1 ^ H 2 | E1 ,..., E l ) P ( E1 ,... E l ) P ( H1 ^ H 2 ) P ( H1 ) P ( H 2 ) because they are independent P ( E1 ,...El | H1 ^ H 2 ) lj 1 P ( E j | H1 ^ H 2 ) assuming E j are independent, given H1 ^ H 2 How to compute P ( E j | H1 ^ H 2 ) ? 18 – Assuming H1 ,..., H n are independent, given E1 ,..., E l ) ? P ( H1 ^ H 2 | E1 ,..., E l ) P ( H1 | E1 ,..., E l ) P ( H 2 | E1 ,..., E l ) but this is a very unreasonable assumption E: earth quake B: burglar E and B are independent But when A is given, they are (adversely) dependent A: alarm set off because they become competitors to explain A • Cannot handle causal chaining P(B|A,E) <<P(B|A) – Ex. A: weather of the year B: cotton production of the year C: cotton price of next year – Observed: A influences C – The influence is not direct (A –> B –> C) P(C|B, A) = P(C|B): instantiation of B blocks influence of A on C • Need a better representation and a better assumption 19 Bayesian Networks (BNs) • Definition: BN = (DAG, CPD) – DAG: directed acyclic graph (BN’s structure) • Nodes: random variables (typically binary or discrete, but methods also exist to handle continuous variables) • Arcs: indicate probabilistic dependencies between nodes (lack of link signifies conditional independence) – CPD: conditional probability distribution (BN’s parameters) • Conditional probabilities at each node, usually stored as a table (conditional probability table, or CPT) P ( xi | i ) where i is the set of all parent nodes of xi – Root nodes are a special case – no parents, so just use priors in CPD: i , so P ( xi | i ) P ( xi ) 20 Example BN P(a0) = 0.001 A P(b0|a0) = 0.3 P(c0|a0) = 0.2 P(b0|a1) = 0.001 B C P(c0|a0) = 0.005 D E P(d0|…) b0 b1 P(d0|b0, c0) = 0.1 P(d0|b0, c1) = 0.01 P(e0|c0) = 0.4 c0 0.1 0.01 P(d0|b1, c0) = 0.01 P(e0|c1) = 0.002 c1 0.01 0.00001 P(d0|b1, c1) = 0.00001 Uppercase: variables (A, B, …) Lowercase: values/states of variables (A has two states a0 and a1) Note that we only specify P(a0) etc., not P(a1), since they have to add to one 21 Netica • An commercial BN package by Norsys • Down load limited version for free from http://www.norsys.com/ • May also down load APIs 22 Conditional independence and chaining • Conditional independence assumption – P ( xi | i , q) P ( xi | i ) i where q is any set of variables q (nodes) other than x i and its successors xi – i blocks influence of other nodes on x i and its successors (q influences x i only through variables in i ) – With this assumption, the complete joint probability distribution of all variables in the network can be represented by (recovered from) local CPDs by chaining these CPDs: P ( x1 ,..., x n ) n1 P ( x i | i ) i 23 Chaining: Example A Computing the joint probability for all variables is easy: B C The joint distribution of all variables D E P(A, B, C, D, E) = P(E | A, B, C, D) P(A, B, C, D) by Bayes’ theorem = P(E | C) P(A, B, C, D) by cond. indep. assumption = P(E | C) P(D | A, B, C) P(A, B, C) = P(E | C) P(D | B, C) P(C | A, B) P(A, B) = P(E | C) P(D | B, C) P(C | A) P(B | A) P(A) For a particular state: P(a0, b0, c1, d1, e0) = P(a0)P(b0|a0)P(c1|a0)P(d1|b0, c1)P(e0| c1) = 0.001*0.3*0.8*0.99*0.002 = 4.752*10^(-7) 24 P(E) = 0.002 P(B) = 0.01 E: earth quake B: burglar A: alarm set off P(E|A) = 0.167; P(B|A) = 0.835 P(A|…) B ~B E 0.9 0.8 P(E|A, E) = 1.0; P(B|A, E) = 0.0112 ~E 0.8 0.0 P(B|A, E) = P(B,A,E)/P(A,E) = P(B,A,E)/(P(B,A,E) + P(~B,A,E) = 0.01*0.002*0.9/(0.01*0.002*0.9 + 0.99*0.002*0.8) = 0.000018/(0.000018 + 0.001548) = 0.000018/0.001566 = 0.01123 25 Topological semantics • A node is conditionally independent of its non- descendants given its parents • A node is conditionally independent of all other nodes in the network given its parents, children, and children’s parents (also known as its Markov blanket) • The method called d-separation can be applied to decide whether a set of nodes X is independent of another set Y, given a third set Z A A B C B A B C C Chain: A and C Diverging: B and Converging: B and are independent, C are independent, C are independent, given B given A NOT given A 26 Inference tasks • Simple queries: Computer posterior probability P(Xi | E=e) – E.g., P(NoGas | Gauge=empty, Lights=on, Starts=false) – Posteriors for ALL nonevidence nodes – Priors for all nodes (E = ) • Conjunctive queries: – P(Xi, Xj | E=e) = P(Xi | E=e) P(Xj | Xi, E=e) • Optimal decisions: Decision networks or influence diagrams – include utility information and actions; – inference is to find P(outcome | action, evidence) • Value of information: Which evidence should we seek next? • Sensitivity analysis: Which probability values are most critical? • Explanation: Why do I need a new starter motor? 27 • MAP problems (explanation) – Let X denote the set of all variables in a BN, V X the set of instantiat ed variables , U X V the set of all un - instantiat ed varialbes. Then the MAP (maximum aposteriori probabilit y) problem is to find the most probable instantiat ion of U , given V , i.e., max u ( P(U | V )) – The solution provides a good explanation for your action – This is an optimization problem 28 Approaches to inference • Exact inference – Enumeration – Variable elimination – Belief propagation in polytrees (singly connected BNs) – Clustering / junction tree algorithms • Approximate inference – Stochastic simulation / sampling methods • Markov chain Monte Carlo methods – Loopy propagation – Others • Mean field theory • Neural networks 29 Inference by enumeration • To compute P(X|E=e), where X is a single variable and E is evidence (instantiation of a set of variables) • Add all of the terms (atomic event probabilities) from the full joint distribution that are consistent with E • If Y are the other (unobserved) variables, excluding X, then the posterior distribution P(X|E=e) = α P(X, e) = α ∑yP(X, e, Y) • Sum is over all possible instantiations of variables in Y • Each P(X, e, Y) term can be computed using the chain rule • Computationally expensive! 30 A Example: Enumeration B C D E • P(xi) = Σ πi P(xi | πi) P(πi) • Suppose we want P(D), and only the value of E is given as true • P (D|e) = ΣA,B,CP(a, b, c, d, e) = ΣA,B,CP(a) P(b|a) P(c|a) P(d|b,c) P(e|c) • With simple iteration to compute this expression, there’s going to be a lot of repetition (e.g., P(e|c) has to be recomputed every time we iterate over C for all possible assignments of A and B)) 31 Exercise: Enumeration p(smart)=.8 p(study)=.6 smart study p(fair)=.9 prepared fair p(prep|…) smart smart study .9 .7 pass study .5 .1 smart smart p(pass|…) prep prep prep prep Query: What is the fair .9 .7 .7 .2 probability that a student studied, given that they pass fair .1 .1 .1 .1 the exam? 32 Variable elimination • Basically just enumeration, but with caching of local calculations • Linear for polytrees • Potentially exponential for multiply connected BNs Exact inference in Bayesian networks is NP-hard! 33 Variable elimination General idea: • Write query in the form P( X n , e ) L P( x | pa )i i xk x3 x2 i • Iteratively – Move all irrelevant terms outside of innermost sum – Perform innermost sum, getting a new term – Insert the new term into the product 34 Variable elimination 8 x 4 = 32 multiplications Example: ΣAΣBΣCP(a) P(b|a) P(c|a) P(d|b,c) P(e|c) = ΣAΣBP(a)P(b|a)ΣCP(c|a) P(d|b,c) P(e|c) 8 x 2 + 4 + 2 = 22 = ΣAP(a)ΣBP(b|a)ΣCP(c|a) P(d|b,c) P(e|c) multiplications for each state of A = a Variable C is summed out for each state of B = b compute fC(a, b) = ΣCP(c|a) P(d|b,c) P(e|c) compute fB(a) = ΣBP(b)fC(a, b) Compute result = ΣAP(a)fB(a) variable B is summed out Here fC(a, b), fB(a) are called factors, which are vectors or matrices 35 Exercise: Variable elimination p(smart)=.8 p(study)=.6 smart study p(fair)=.9 prepared fair p(prep|…) smart smart study .9 .7 pass study .5 .1 smart smart p(pass|…) prep prep prep prep Query: What is the fair .9 .7 .7 .2 probability that a student is smart, given that they pass fair .1 .1 .1 .1 the exam? 36 Belief Propagation • Singly connected network, SCN (also known as polytree) – there is at most one undirected path between any two nodes (i.e., the network is a tree if the direction of arcs are ignored) – The influence of the instantiated variable (evidence) spreads to the rest of the network along the arcs • The instantiated variable influences its predecessors and successors differently (using CPT along opposite A directions) • Computation is linear to the diameter of B C the network (the longest undirected path) D E=e F • Update belief (posterior) of every non- evidence node in one pass – For multi-connected net: conditioning 37 Conditioning A B C D E • Conditioning: Find the network’s smallest cutset S (a set of nodes whose removal renders the network singly connected) – In this network, S = {A} or {B} or {C} or {D} • For each instantiation of S, compute the belief update with the belief propagation algorithm • Combine the results from all instantiations of S (each is weighted by P(S = s)) • Computationally expensive (finding the smallest cutset is in general NP-hard, and the total number of possible instantiations of S is O(2|S|)) 38 Junction Tree • Convert a BN to a junction tree – Moralization: add undirected edge between every pair of parents, then drop directions of all arc: Moralized Graph – Triangulation: add an edge to any cycle of length > 3: Triangulated Graph – A junction tree is a tree of cliques of the triangulated graph – Cliques are connected by links • A link stands for the set of all variables S shared by these two cliques • Each clique has a potential (similar to CPT), constructed from CPT of variables in the original BN 39 Junction Tree • Example A A B C B C A,B,C D E D E Triangulated graph (B, C) A simple BN A B,C,D (C, D) B C C,D.E D E Junction tree of 3 nodes Moralized graph 40 Junction Tree • Reasoning – Since it is now a tree, polytree algorithm can be applied, but now two cliques exchange P(S), the distribution over S, their shared variables. – Complexity: • O(n) steps, where n is the number of cliques • Each step is expensive if cliques are large (CPT exponential to clique size) • Construction of CPT of JT is expensive as well, but it needs to compute only once. 41 Some comments on BN reasoning – Let X ( X 1 , K , X n ) be the set of all variables in a BN. Any BN reasoning task can be expressed in the form of calculating P(U | V ) where U ,V X – This can be done by marginalization of the joint distribution P(X) over Y = X \ U \ V: Y P(U ,V , Y ) where each entry P(x) = P(u,v,y) can be calculated by chain rule from CPTs – Computation can be done more efficiently using, say Junction tree, by utilizing variable interdependencies – Computational complexity of BN reasoning is proved to be NP- hard by reducing 3SAT problems to BN reasoning (Cooper 1990) 42 Approximate inference: Direct sampling • Suppose you are given values for some subset of the variables, E, and want to infer distributions for unknown variables, Z • Randomly generate a very large number of instantiations from the BN according to the distribution – Generate instantiations for all variables – start at root variables and work your way “forward” in topological order • Rejection sampling: Only keep those instantiations that are consistent with the values for E • Use the frequency of values for Z to get estimated probabilities • Accuracy of the results depends on the size of the sample (asymptotically approaches exact results) • Very expensive and inefficient 43 Likelihood weighting • Idea: Don’t generate samples that need to be rejected in the first place! • Sample only from the unknown variables Z and X (E are fixed) • Weight each sample according to the likelihood that it would occur, given the evidence E – A weight w is associated with each sample (w initialized to 1) – When a evidence node (say E1 = e1-0) is selected for weighting, its parents are already instantiated (say parents A and B are assigned state a and b) – Modify w = w * P(e1-0 | a, b) based on E1’s CPT – Repeat for the other evidence nodes 44 Markov chain Monte Carlo algorithm • So called because – Markov chain – each instance generated in the sample is dependent on the previous instance – Monte Carlo – statistical sampling method • Perform a random walk through variable assignment space, collecting statistics as you go – Start with a random instantiation, consistent with evidence variables – At each step, randomly select a non-evidence variable x, randomly sample its value by P( x | mb( x)) P( x | parent( x)) Π ( P( y | parentsY ) Ychild ( X ) • Given enough samples, MCMC gives an accurate estimate of the true distribution of values 45 Loopy Propagation • Belief propagation – Works only for polytrees (exact solution) – Each evidence propagates once throughout the network • Loopy propagation – Let propagation continue until the network stabilize (hope) • Experiments show – Many BN stabilize with loopy propagation – If it stabilizes, often yielding exact or very good approximate solutions • Analysis – Conditions for convergence and quality approximation are under intense investigation 46 Noisy-Or BN • A special BN of binary variables (Peng & Reggia, Cooper) – Each link xi x j is associated with a probabilit y value called causal strength cij that measures the strength of xi alone may cause x j , i.e., cij P( xi | x j is true and all others in i are false) – Causation independence: parent nodes influence a child independently • Advantages: – One-to-one correspondence between causal links and causal strengths – Easy for humans to understand (acquire and evaluate KB) – Fewer # of probabilities needed in KB Complete joint prob. distributi on : 2 n General BN : in1 2|i | Noisy - Or BN : in1| i | – Computation is less expensive • Disadvantage: less expressive (less general) 47 Learning BN (from case data) • Needs for learning – Difficult to construct BN by humans (esp. CPT) – Experts’ opinions are often biased, inaccurate, and incomplete – Large databases of cases become available • What to learn – Parameter learning: learning CPT when DAG is known (easy) – Structural learning: learning DAG (hard) • Difficulties in learning DAG from case data – There are too many possible DAG when # of variables is large (more than exponential) n # of possible DAG 3 25 10 4*10^18 – Missing values in database – Noisy data 48 BN Learning Approaches • Early effort: Based on variable dependencies (Pearl) – Find all pairs of variables that are dependent of each other (applying standard statistical method on the database) – Eliminate (as much as possible) indirect dependencies – Determine directions of dependencies – Learning results are often incomplete (learned BN contains indirect dependencies and undirected links) 49 BN Learning Approaches • Bayesian approach (Cooper) – Find the most probable DAG, given database DB, i.e., max(P(DAG|DB)) or max(P(DAG, DB)) – Based on some assumptions, a formula is developed to compute P(DAG, DB) for a given pair of DAG and DB – A hill-climbing algorithm (K2) is developed to search a (sub)optimal DAG – Compute CPTs after the DAG is determined – Extensions to handle some form of missing values 50 BN Learning Approaches • Minimum description length (MDL) (Lam, etc.) – Sacrifices accuracy for simpler (less dense) structure • Case data not always accurate • Outliers are hard to model (needs more links) • Fewer links imply smaller CPD tables and less expensive inference – L = L1 + L2 where • L1: the length of the encoding of DAG (smaller for simpler DAG) • L2: the length of the encoding of the difference between DAG and DB (smaller for better match of DAG with DB) • Smaller L1 implies less accurate DAG, and thus larger L2 – Find DAG by heuristic best-first search, that Minimizes L 51 BN Learning Approaches • Neural network approach (Neal, Peng) – For noisy-or BN ~ Maximizing L ln P(V V r ) where V r D D : case database; V r : case in D; ~ V : state vector of the learned network L measures the similarity of the two distributi ons : one in D, another in the learned network – Change inter-node link strength locally, following gradient descent approach to maximize L. 52 • Compare Neural network approach with Cooper’s K2 • Network: Alarm (37 nodes) # cases missing links extra links time 500 2/0 2/6 63.76/5.91 1000 0/0 1/1 69.62/6.04 2000 0/0 0/0 77.45/5.86 10000 0/0 0/0 161.97/5.83 53 Current research in BN • Missing data – Missing value: EM (expectation maximization) – Missing (hidden) variables are harder to handle • BN with time – Dynamic BN: assuming temporal relation obey Markov chain • Cyclic relations – Often found in social-economic analysis – Using dynamic BN? • Continuous variable – Some work on variables obeying Gaussian distribution • Connecting to other fields – Databases; Statistics; Symbolic AI (FOL); Semantic web • Reasoning with uncertain evidence – Virtual evidence – Soft evidence 54 Other formalisms for Uncertainty Fuzzy sets and fuzzy logic • Ordinary set theory – f A ( x) 1 if x A 0 otherwise f A ( x) is called the characteri stic or membership function of set A 1 if x A Predicate A( x) 0 otherwise When it is uncertain if x A , use probabilit y P ( x A ) – There are sets that are described by vague linguistic terms (sets without hard, clearly defined boundaries), e.g., tall-person, fast- car • Continuous • Subjective (context dependent) • Hard to define a clear-cut 0/1 membership function 55 • Fuzzy set theory – Relax f A ( x ) from binary {0,1} to continuous[0,1] o stands for thedegree x is thought t belong to set A height(john) = 6’5” Tall(john) = 0.9 height(harry) = 5’8” Tall(harry) = 0.5 height(joe) = 5’1” Tall(joe) = 0.1 – Examples of membership functions 1- Set of teenagers 0 12 19 1- Set of young people 0 12 19 1- Set of mid-age people 20 35 50 65 80 56 • Fuzzy logic: many-value logic – Fuzzy predicates (degree of truth) FA ( x) y if f A ( x) y – Connectors/Operators negation : FA ( x) 1 FA ( x) conjunction : FA B ( x) min{FA ( x) , FB ( x)} disjunction : FA B ( x) max{ FA ( x) , FB ( x)} • Compare with probability theory – Prob. Uncertainty of outcome, • Based on large # of repetitions or instances • For each experiment (instance), the outcome is either true or false (without uncertainty or ambiguity) unsure before it happens but sure after it happens Fuzzy: vagueness of conceptual/linguistic characteristics • Unsure even after it happens whether a child of tall mother and short father is tall unsure before the child is born unsure after grown up (height = 5’6”) 57 – Empirical vs subjective (testable vs agreeable) – Fuzzy set connectors may lead to unreasonable results • Consider two events A and B with P(A) < P(B) • If A => B (or A B) then P(A ^ B) = P(A) = min{P(A), P(B)} P(A v B) = P(B) = max{P(A), P(B)} • Not the case in general P(A ^ B) = P(A)P(B|A) P(A) P(A v B) = P(A) + P(B) – P(A ^ B) P(B) (equality holds only if P(B|A) = 1, i.e., A => B) – Something prob. theory cannot represent • Tall(john) = 0.9, ~Tall(john) = 0.1 Tall(john) ^ ~Tall(john) = min{0.1, 0.9) = 0.1 john’s degree of membership in the fuzzy set of “median- height people” (both Tall and not-Tall) • In prob. theory: P(john Tall ^ john Tall) = 0 58 Uncertainty in rule-based systems • Elements in Working Memory (WM) may be uncertain because – Case input (initial elements in WM) may be uncertain Ex: the CD-Drive does not work 70% of the time – Decision from a rule application may be uncertain even if the rule’s conditions are met by WM with certainty Ex: flu => sore throat with high probability • Combining symbolic rules with numeric uncertainty: Mycin’s Certainty Factor (CF) – An early attempt to incorporate uncertainty into KB systems – CF [-1, 1] – Each element in WM is associated with a CF: certainty of that assertion – Each rule C1,...,Cn => Conclusion is associated with a CF: certainty of the association (between C1,...Cn and Conclusion). 59 – CF propagation: • Within a rule: each Ci has CFi, then the certainty of Action is min{CF1,...CFn} * CF-of-the-rule • When more than one rules can apply to the current WM for the same Conclusion with different CFs, the largest of these CFs will be assigned as the CF for Conclusion • Similar to fuzzy rule for conjunctions and disjunctions – Good things of Mycin’s CF method • Easy to use • CF operations are reasonable in many applications • Probably the only method for uncertainty used in real-world rule-base systems – Limitations • It is in essence an ad hoc method (it can be viewed as a probabilistic inference system with some strong, sometimes unreasonable assumptions) • May produce counter-intuitive results. 60 Dempster-Shafer theory • A variation of Bayes’ theorem to represent ignorance • Uncertainty and ignorance – Suppose two events A and B are ME and EXH, given an evidence E A: having cancer B: not having cancer E: smoking – By Bayes’ theorem: our beliefs on A and B, given E, are measured by P(A|E) and P(B|E), and P(A|E) + P(B|E) = 1 – In reality, I may have some belief in A, given E I may have some belief in B, given E I may have some belief not committed to either one, – The uncommitted belief (ignorance) should not be given to either A or B, even though I know one of the two must be true, but rather it should be given to “A or B”, denoted {A, B} – Uncommitted belief may be given to A and B when new evidence is discovered 61 • Representing ignorance – Frame of discernment :q {h1 ,...,hn }, a set of ME and EXH hypotheses. The power set 2q is organized as a lattice of super/subs et relations. Each node S is a subset of hypotheses( S q ) – Ex: q = {A,B,C} Each node S is associated with a {A,B,C} 0.15 basic probabilit y assignment m ( S ) 0 m ( S ) 1; {A,B} 0.1 {A,C} 0.1 {B,C}0.05 m () 0; Sq m(S) 1 {A} 0.1 {B} 0.2 {C}0.3 • Belief function {} 0 Bel ( S ) S ' S m ( S ' ); Bel () 0; Bel (q ) 1 Bel ({A, B}) m ({A, B}) m ({A}) m ({B}) m () 0.1 0.1 0.2 0 0.4 Bel ({A, B}C ) Bel ({C}) 0.3 62 – Plausibility (upper bound of belief of a node) All belief not committed to S C may be commited to S Pls( S ) 1 Bel ( S C ) Pls({A, B}) 1 Bel ({C}) 1 0.3 0.7 [ Bel ( S ), Pls( S )] belief interval Lower Upper {A,B,C} 0.15 bound bound (known (maximally {A,B} 0.1 {A,C} 0.1 {B,C}0.05 belief) possible) {A} 0.1 {B} 0.2 {C}0.3 {} 0 63 • Evidence combination (how to use D-S theory) – Each piece of evidence has its own m(.) function for the same q q { A, B} : A : having cancer; B : not having cancer {A,B} 0.3 {A,B} 0.1 {A} 0.2 {B} 0.5 {A} 0.7 {B} 0.2 {} 0 {} 0 m1 ( S ) m2 ( S ) E1 : smoking E2 : living in high radiation area – Belief based on combined evidence can be computed from m( S ) m1 ( S ) m2 ( S ) X Y S m1 ( X )m2 (Y ) 1 X Y m1 ( X )m2 (Y ) normalization factor incompatible combination 64 {A,B} 0.3 {A,B} 0.1 {A,B} 0.049 {A} 0.2 {B} 0.5 {A} 0.7 {B} 0.2 {A} 0.607 {B} 0.344 {} 0 {} 0 {} 0 E1 E2 E1 ^ E2 m1 ({A})m 2 ({A}) m1 ({A})m 2 ({A, B}) m1 ({A, B})m 2 ({A}) m ({A}) 1 [m1 ({A})m 2 ({B}) m1 ({B})m 2 ({A})] 0.2 0.7 0.2 0.1 0.3 0.7 0.37 0.607 1 [0.2 0.2 0.5 0.7] 0.61 m1 ({B})m 2 ({B}) m1 ({B})m 2 ({A, B}) m1 ({A, B})m 2 ({B}) m ({B}) 1 [m1 ({A})m 2 ({B}) m1 ({B})m 2 ({A})] 0.5 0.2 0.5 0.1 0.3 0.2 0.21 0.344 1 [0.2 0.2 0.5 0.7] 0.61 m1 ({A, B})m2 ({A, B}) 0.03 m({A, B}) 0.049 0.61 0.61 65 – Ignorance is reduced from m1({A,B}) = 0.3 to m({A,B}) = 0.049) – Belief interval is narrowed A: from [0.2, 0.5] to [0.607, 0.656] B: from [0.5, 0.8] to [0.344, 0.393] • Advantage: – The only formal theory about ignorance – Disciplined way to handle evidence combination • Disadvantages – Computationally very expensive (lattice size 2^|q|) – Assuming hypotheses are ME and EXH – How to obtain m(.) for each piece of evidence is not clear, except subjectively 66