Upload
nouran-m-radwan
View
26
Download
4
Embed Size (px)
Citation preview
Agenda Fuzzy logic Fuzzy Sets Crisp and Fuzzy Sets Experts are vague Fuzzy Expert Systems Fuzzy Rules
Fuzzy Logic Applications
Fuzzy Expert System Applications
• Journals of Fuzzy to find Researches
Fuzzy Logic Fuzzy logic is based on the idea of Fuzzy logic is based on the idea of varying
degrees of truth. . Computers can apply this logic to
represent vague and imprecise ideas, such as ttemperature, height, speed, distance, emperature, height, speed, distance, beauty beauty all come on a sliding scale. all come on a sliding scale.
The motor is running The motor is running really hotreally hot. .
Fuzz
y Se
tsFu
zzy
Sets
The basic idea of the fuzzy set theory is that an element belongs to a fuzzy set with a certain degree of membership.
Crisp and Fuzzy SetsCrisp and Fuzzy Sets
150 210170 180 190 200160Height, cm
Degree ofMembership
Tall Men
150 210180 190 200
1.0
0.0
0.2
0.4
0.6
0.8
160
Degree ofMembership
Short Average ShortTall
170
1.0
0.0
0.2
0.4
0.6
0.8
Fuzzy Sets
Crisp Sets
Short Average
Tall
Tall
Tall men Tall men ={0/180, ={0/180, 0.5/185, 1/190}0.5/185, 1/190}
Average men Average men ={0/165, 1/175, ={0/165, 1/175, 0/185}0/185}
Short men Short men ={1/160, ={1/160, 0.5/165, 0/1700.5/165, 0/170}}
A man who is 184 cm tall is a member of the average men set with a degree of membership of 0.1, and at the same time, he is also a member of the tall men set with a degree of 0.4.
To design an expert system a major task is to codify To design an expert system a major task is to codify the expert’s decision-making process. the expert’s decision-making process.
In a domain there may be precise, scientific tests In a domain there may be precise, scientific tests and measurements. Some of the decisions could be and measurements. Some of the decisions could be done using traditional logic.done using traditional logic.
But others are Fuzzy. Fuzzy systems afford a But others are Fuzzy. Fuzzy systems afford a broader, richer field of data and manipulations than broader, richer field of data and manipulations than do more traditional methods.do more traditional methods.
Fuzzy Expert SystemFuzzy Expert System
A fuzzy expert system is an expert system that uses fuzzy logic instead of crisp logic.
A fuzzy expert system is a collection of membership functions and rules that are used to reason about data.
Building a Fuzzy Expert systemBuilding a Fuzzy Expert system
1.1. Specify the problem and define linguistic Specify the problem and define linguistic variables.variables.
2.2. Determine fuzzy sets and Membership function.Determine fuzzy sets and Membership function.3.3. Elicit and Construct Fuzzy rules.Elicit and Construct Fuzzy rules.4.4. Encode the fuzzy sets, fuzzy rules and Encode the fuzzy sets, fuzzy rules and
procedures to perform fuzzy inference into the procedures to perform fuzzy inference into the expert system.expert system.
5.5. Evaluate and tune the system.Evaluate and tune the system.
Fuzzy RulesFuzzy Rules
In 1973, Lotfi Zadeh presented a new In 1973, Lotfi Zadeh presented a new approach to analysis of complex systems.approach to analysis of complex systems.
He suggested capturing human He suggested capturing human knowledge in fuzzy rules.knowledge in fuzzy rules.
What is the difference between Crisp and Fuzzy
rules?
Crisp IF-THEN rule Crisp IF-THEN rule Rule: 1Rule: 1IF speed is > 100 THEN stopping_distance is longIF speed is > 100 THEN stopping_distance is longRule: 2Rule: 2IF speed is < 40 THEN stopping_distance is shortIF speed is < 40 THEN stopping_distance is short Fuzzy IF-THEN ruleFuzzy IF-THEN ruleRule: 1Rule: 1IF speed is IF speed is fastfast THEN stopping_distance is long THEN stopping_distance is longRule: 2Rule: 2IF speed is IF speed is slow slow THEN stopping_distance is shortTHEN stopping_distance is short
Fuzzy RulesFuzzy Rules
In fuzzy expert systems, linguistic variables are used in fuzzy rules
IF wind is strong THEN sailing is good. IF project duration is long THEN
completion risk is high. IF speed is slow THEN stopping
distance is short.
A fuzzy rule can have multiple antecedentsA fuzzy rule can have multiple antecedentsIF project_duration is long AND project_staffing IF project_duration is long AND project_staffing is large AND project_funding is inadequate THEN is large AND project_funding is inadequate THEN risk is high.risk is high.IFIF service is excellent service is excellent OR food is delicious OR food is delicious THEN tip is generous. THEN tip is generous. The consequent can include multiple partsThe consequent can include multiple partsIF temperature is hot THEN hot_water is IF temperature is hot THEN hot_water is reduced; cold_water is increased.reduced; cold_water is increased.
Fuzzy RulesFuzzy Rules
Fuzzy Expert SystemFuzzy Expert System
Crisp Input
Fuzzy Input
Fuzzy Output
Crisp Output
Fuzzification
Rule Evaluation
Defuzzification
Input Membership Functions
Rules / Inferences
Output Membership Functions
Fuzzification
In the process of fuzzification,
Membership functions are defined. Mapping the crisp inputs to fuzzy set values from 0 to 1 using a set of input membership functions.
Inference
In the process of inference, Degree of Support of each rule is computed. Fuzzy inference is the process of applying reasoning to compute fuzzy outputs. Fuzzy inference is a decision making unit where testing of all of the rules in a fuzzy rules are performed and integrated to make a decision.
Defuzzification
In the process of defuzzification,
Convert the fuzzy value obtained from composition into a “crisp” value. This process is often complex since the fuzzy set might not translate directly into a crisp value.
Fuzzy Logic Applications• Aerospace
– Control of spacecraft, satellite and aircraft.• Defense
– Automatic target recognition of thermal infrared images, naval decision support aids, fuzzy set modeling of NATO decision making.
• Electronics– Control of automatic exposure in video cameras, air
conditioning systems, microwave ovens.
• Marine– Autopilot for ships, optimal route selection, control
of autonomous underwater vehicles, ship steering.• Robotics
– Fuzzy control for flexible-link manipulators, robot arm control.
• Securities– Decision systems for securities trading.
Fuzzy Logic Applications
• Business– Decision-making support systems, personnel
evaluation in a large company.• Financial
– Banknote transfer control, fund management, stock market predictions.
• Manufacturing– Optimization of cheese production.
Fuzzy Logic Applications
• Industrial– Quantitative pattern analysis for industrial quality
assurance, control of water purification plants.• Medical
– Medical diagnostic support system, control of arterial pressure during anesthesia, modeling of neuropathological findings in Alzheimer's patients, radiology diagnoses.
Fuzzy Logic Applications
Fuzzy Expert System Fuzzy Expert System Applications Applications
Human Disease Diagnosis Using a Fuzzy Expert Human Disease Diagnosis Using a Fuzzy Expert System [2010]System [2010]
Mir Anamul Hasan et al. make a comparative Mir Anamul Hasan et al. make a comparative analysis to identify which symptoms are major analysis to identify which symptoms are major symptoms for particular diseases, which will be symptoms for particular diseases, which will be used to diagnosis by the fuzzy expert system. used to diagnosis by the fuzzy expert system.
User has to answer some question that based User has to answer some question that based on the knowledge on the basis of IF-THEN rule on the knowledge on the basis of IF-THEN rule of the expert system.of the expert system.
Fuzzy Expert System Fuzzy Expert System Applications Applications
Application of Expert System with Fuzzy Application of Expert System with Fuzzy Logic in Teachers Performance Evaluation Logic in Teachers Performance Evaluation [2011][2011]
Abdur Rashid Khan et al. extract a set of 99 Abdur Rashid Khan et al. extract a set of 99 attributes that are ranked according to it’s attributes that are ranked according to it’s effect on the teacher performance.effect on the teacher performance.
By using IF-THEN rule the teachers By using IF-THEN rule the teachers performance is either ranked high, very high performance is either ranked high, very high etc. etc.
Fuzzy Expert System Fuzzy Expert System Applications Applications
Fuzzy Expert System for Diabetes using Fuzzy Expert System for Diabetes using Fuzzy Verdict Mechanism [2011]Fuzzy Verdict Mechanism [2011]
M. Kalpana et al. select five attributes M. Kalpana et al. select five attributes according to the American Diabetes according to the American Diabetes Association & OGTT five attributes as the Association & OGTT five attributes as the input fuzzy variables of the adopted fuzzy input fuzzy variables of the adopted fuzzy rule-based inference system. rule-based inference system.
The proposed fuzzy expert system The proposed fuzzy expert system implemented with the MATLAB.implemented with the MATLAB.
Fuzzy Expert System Fuzzy Expert System Applications Applications
Fuzzy Knowledge based for Tropical Infectious Fuzzy Knowledge based for Tropical Infectious Disease Diagnosis[2012]Disease Diagnosis[2012]
Putu Manik Prihatini et al. make a knowledge Putu Manik Prihatini et al. make a knowledge acquisition which is obtained for the seven acquisition which is obtained for the seven diseases, and 22 symptoms. diseases, and 22 symptoms.
They use SQLyog community edition to build They use SQLyog community edition to build knowledge base, Macromedia Dreamweaver knowledge base, Macromedia Dreamweaver 8 with PHP and java script to build the 8 with PHP and java script to build the application and CSS to design the interface.application and CSS to design the interface.
Fuzzy Expert System Fuzzy Expert System Applications Applications
A Novel Web-based Human Advisor Fuzzy A Novel Web-based Human Advisor Fuzzy Expert System [2013] Expert System [2013] Vahid Rafe and Mahdi.Vahid Rafe and Mahdi.
User selects the type of advisory service and User selects the type of advisory service and enters crisp data, then system asks the question enters crisp data, then system asks the question related to the problem. related to the problem.
CHA translates the user input to linguistic CHA translates the user input to linguistic variables, makes the fuzzy rule and generate the variables, makes the fuzzy rule and generate the fuzzy answer then fuzzy answer are deffuzzified fuzzy answer then fuzzy answer are deffuzzified to crisp output and is reported to user. to crisp output and is reported to user.
Advances in Fuzzy Systems• http://www.hindawi.com/journals/afs/2014/ Advances in Fuzzy Systems is an
international journal which aims to help promote the advances in the development and practice of fuzzy system technologies in the areas of engineering, management, medical, economic, environmental, and societal problems.
Fuzzy Sets and Systemshttp://www.journals.elsevier.com/fuzzy-sets-and-systems/
Fuzzy Information and Engineeringhttp://www.journals.elsevier.com/fuzzy-information-and-engineering/
Journal Intelligent and Fuzzy Systems
http://www.iospress.nl/journal/journal-of-intelligent-fuzzy-systems/