Bayesian Decision Theory Compiled By: Raj Gaurang Tiwari Assistant Professor SRMGPC, Lucknow

Statistical Pattern Recognition The design of a recognition system requires careful attention to the following issues: definition of pattern classes, sensing environment pattern representation feature extraction and selection cluster analysis classifier design and learning selection of training and test samples performance evaluation.

Statistical Pattern Recognition….. In statistical pattern recognition, a pattern is represented by a set of d features, or attributes, viewed as a d-dimensional feature vector. Well-known concepts from statistical decision theory are utilized to establish decision boundaries between pattern classes. The recognition system is operated in two modes: training (learning) and classification (testing)

Model for statistical pattern recognition

The role of the preprocessing module is to segment the pattern of interest from the background, remove noise, normalize the pattern, and any other operation which will contribute in defining a compact representation of the pattern. In the training mode, the feature extraction/selection module finds the appropriate features for representing the input patterns and the classifier is trained to partition the feature space. The feedback path allows a designer to optimize the preprocessing and feature extraction/selection strategies. In the classification mode, the trained classifier assigns the input pattern to one of the pattern classes under consideration based on the measured features.

Decision theory Decision theory is the study of making decisions that have a significant impact Decision-making is distinguished into: Decision-making under certainty Decision-making under non-certainty Decision-making under risk Decision-making under uncertainty

Probability theory Most decisions have to be taken in the presence of uncertainty Probability theory quantifies uncertainty regarding the occurrence of events or states of the world Basic elements of probability theory: Random variables describe aspects of the world whose state is initially unknown Each random variable has a domain of values that it can take on (discrete, boolean, continuous) An atomic event is a complete specification of the state of the world, i.e. an assignment of values to variables of which the world is composed

Probability Theory.. Probability space Axioms The sample space S={e1 ,e2 ,…,en } which is a set of atomic events Probability measure P which assigns a real number between 0 and 1 to the members of the sample space Axioms All probabilities are between 0 and 1 The sum of probabilities for the atomic events of a probability space must sum up to 1 The certain event S (the sample space itself) has probability 1,and the impossible event which never occurs, probability 0

Prior Priori Probabilities or Prior reflects our prior knowledge of how likely an event occurs. In the absence of any other information, a random variable is assigned a degree of belief called unconditional or prior probability

Class Conditional probability When we have information concerning previously unknown random variables then we use posterior or conditional probabilities: P(a|b) the probability of a given event a that we know b Alternatively this can be written (the product rule): P(a b)=P(a|b)P(b)

Bayes’ rule The product rule can be written as: P(a b)=P(a|b)P(b) P(a b)=P(b|a)P(a) By equating the right-hand sides: This is known as Bayes’ rule

Bayesian Decision Theory Bayesian Decision Theory is a fundamental statistical approach that quantifies the tradeoffs between various decisions using probabilities and costs that accompany such decisions. Example: Patient has trouble breathing – Decision: Asthma versus Lung cancer – Decide lung cancer when person has asthma Cost: moderately high (e.g., order unnecessary tests, scare patient) – Decide asthma when person has lung cancer Cost: very high (e.g., lose opportunity to treat cancer at early stage, death)

Decision Rules Progression of decision rules: – (1) Decide based on prior probabilities – (2) Decide based on posterior probabilities – (3) Decide based on risk

Fish Sorting Example Revisited

Decision based on prior probabilities

Question Consider a two-class problem, { c1 and c2 } where the prior probabilities of the two classes are given by P ( c1 ) = ⋅7 and P ( c2 ) = ⋅3 Design a classification rule for a pattern based only on prior probabilities Calculation of Error Probability – P ( error )

Solution

Decision based on class conditional probabilities

Posterior Probabilities

Bayes Formula Suppose the priors P(wj) and conditional densities p(x|wj) are known, prior likelihood posterior evidence

Making a Decision

Probability of Error Average probability of error P(error) Bayes decision rule minimizes this error because

The dotted line at x0 is a threshold partitioning the feature space into two regions,R1 and R2. According to the Bayes decision rule,for all values of x in R1 the classifier decides 1 and for all values in R2 it decides 2. However, it is obvious from the figure that decision errors are unavoidable. Example of the two regions R1 and R2 formed by the Bayesian classifier for the case of two equiprobable classes. The dotted line at x0 is a threshold partitioning the feature space into two regions,R1 and R2. According to the Bayes decision rule, for all values of x in R1 the classifier decides 1 and for all values in R2 it decides 2. However, it is obvious from the figure that decision errors are unavoidable.

total probability,Pe,of committing a decision error which is equal to the total shaded area under the curves in Figure

Minimizing the Classification Error Probability Show that the Bayesian classifier is optimal with respect to minimizing the classification error probability.

Generalized Bayesian Decision Theory

Bayesian Decision Theory…

Bayesian Decision Theory…

Conditional Risk

Minimum-Risk Classification For every x the decision function α(x) assumes one of the a values α1, ..., αa. The overall risk R is the expected loss associated with a given decision rule.

Two-category classification 1 : deciding 1 2 : deciding 2 ij = (i | j) loss incurred for deciding i when the true state of nature is j Conditional risk: R(1 | x) = 11P(1 | x) + 12P(2 | x) R(2 | x) = 21P(1 | x) + 22P(2 | x)

action 1: “decide 1” is taken Our rule is the following: if R(1 | x) < R(2 | x) action 1: “decide 1” is taken This results in the equivalent rule : decide 1 if: By employingBayes’ formula (21- 11) P(x | 1) P(1) > (12- 22) P(x | 2) P(2) and decide 2 otherwise

Likelihood ratio Then take action 1 (decide 1) Otherwise take action 2 (decide 2)

Example Suppose selection of w1 and w2 has same probability: P(w1)=p(w2)=1/2 Assume that the loss matrix is of the form If misclassification of patterns that come from w2 is considered to have serious consequences, then we must choose 12 > 21.

Thus, patterns are assigned to w2 class if That is, P(x | 1) is multiplied by a factor less than 1

Example

Minimum-Error-Rate Classification The action αi is usually interpreted as the decision that the true state of nature is ωi. Actions are decisions on classes If action i is taken and the true state of nature is j then: the decision is correct if i = j and in error if i  j Seek a decision rule that minimizes the probability of error which is the error rate

Introduction of the zero-one loss function: Therefore, the conditional risk is: “The risk corresponding to this loss function is the average probability error” 

Minimizing the risk requires maximizing P(i | x) (since R(i | x) = 1 – P(i | x)) For Minimum error rate Decide i if P (i | x) > P(j | x) j  i