Fuzzy Inference (Expert) System

Lecture Module 20

Introduction
A Fuzzy Inference System (FIS) is a way of mapping an input space to an output space using fuzzy logic FIS uses a collection of fuzzy membership functions and rules, instead of Boolean logic, to reason about data. The rules in FIS (sometimes may be called as fuzzy expert system) are fuzzy production rules of the form:
if p then q, where p and q are fuzzy statements. For example, in a fuzzy rule if x is low and y is high then z is medium. Here x is low; y is high; z is medium are fuzzy statements; x and y are input variables; z is an output variable, low, high, and medium are fuzzy sets.

Cont
The antecedent describes to what degree the rule applies, while the conclusion assigns a fuzzy function to each of one or more output variables. Most tools for working with fuzzy expert systems allow more than one conclusion per rule. The set of rules in a fuzzy expert system is known as knowledge base. The functional operations in fuzzy expert system proceed in the following steps.
Fuzzification Fuzzy Inferencing (apply implication method) Aggregation of all outputs Defuzzification

Structure of a Fuzzy Expert System

Fuzzification
In the process of fuzzification, membership functions defined on input variables are applied to their actual values so that the degree of truth for each rule premise can be determined. Fuzzy statements in the antecedent are resolved to a degree of membership between 0 and 1.
If there is only one part to the antecedent, then this is the degree of support for the rule. If there are multiple parts to the antecedent, apply fuzzy logic operators and resolve the antecedent to a single number between 0 and 1.

Antecedent may be joined by OR; AND operators.

For OR -- max For AND -- min

Fuzzy Inferencing
In the process of inference
Truth value for the premise of each rule is computed and applied to the conclusion part of each rule. This results in one fuzzy set to be assigned to each output variable for each rule.

The use of degree of support for the entire rule is to shape the output fuzzy set. The consequent of a fuzzy rule assigns an entire fuzzy set to the output. If the antecedent is only partially true, (i.e., is assigned a value less than 1), then the output fuzzy set is truncated according to the implication method.

Cont...
If the consequent of a rule has multiple parts, then all consequents are affected equally by the result of the antecedent. The consequent specifies a fuzzy set to be assigned to the output. The implication function then modifies that fuzzy set to the degree specified by the antecedent. The following functions are used in inference rules. min or prod are commonly used as inference rules.
min: truncates the consequent's membership function prod: scales it.

Aggregation of all outputs

It is the process where the outputs of each rule are combined into a single fuzzy set. The input of the aggregation process is the list of truncated output functions returned by the implication process for each rule. The output of the aggregation process is one fuzzy set for each output variable.
Here, all fuzzy sets assigned to each output variable are combined together to form a single fuzzy set for each output variable using a fuzzy aggregation operator.

Some of the most commonly used aggregation operators are

the maximum : point-wise maximum over all of the fuzzy sets the sum : (point-wise sum over all of the fuzzy the probabilistic sum.

Defuzzification
In Defuzzificztion, the fuzzy output set is converted to a crisp number. Some commonly used techniques are the centroid and maximum methods.
In the centroid method, the crisp value of the output variable is computed by finding the variable value of the centre of gravity of the membership function for the fuzzy value. In the maximum method, one of the variable values at which the fuzzy set has its maximum truth value is chosen as the crisp value for the output variable.

Some other methods for defuzzification are:

bisector, middle of maximum (the average of the maximum value of the output set), largest of maximum, and smallest of maximum, etc.

Generic Method
Main steps are
Evaluate the antecedent for each rule Obtain each rule's conclusion Aggregate conclusions Defuzzification

We will explain these steps using an example of Tipping Problem Two inputs : Quality of food and Service at a restaurant
rated at scale from 0-10

One output: Amount of tip to be given Tip should reflect the quality of the food and service. The tip might be in the range 5-15% of total bill paid.

Rules for Tipping

Let us consider the following three rules

If service is poor or food is bad, then tip is cheap If service is good, then tip is average If service is excellent or food is delicious, then tip is generous

Input variables
Service : represented by poor, good, excellent Food : represented by bad, delicious

Output Variable:
Tip : represented by cheap, average, generous

Antecedent for each rule

Rule's Conclusion

Aggregate Conclusions

Defuzzification

All Steps Together looks like

MatLab

Fuzzy Toolkit

Introduction
MATLAB fuzzy logic toolbox provides facility for the development of fuzzy-logic systems using
graphical user interface (GUI) tools command line functionality

There are five primary GUI tools

Fuzzy Inference System (FIS) Editor Membership Function Editor Rule Editor Rule Viewer Surface Viewer

GUI Tools

Fuzzy Inference System (FIS) Editor

Membership Function Editor

Select & edit attributes of membership function

Display & edit values of current variable

Name & edit parameters of membership function

Rule Editor

Rules automatically updated

Create and edit rules

Rule Viewer

Shows how input variable is used in rules

Shows how output variable is used in rules; shows output of fuzzy system

Surface Viewer

Specify input and output variables

Shows output surface for any system output versus any one (or two) inputs