a formal specification for a fuzzy expert system 1 information and software technology 2003.05 -vol...

20
A formal specification for a fuzzy expert system 1 A formal specification for a fuzzy expert system Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author Chris Matthews speaker hsiu chun-chiu (9254610)

Upload: barrie-thornton

Post on 04-Jan-2016

217 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

1

A formal specification for a fuzzy expert system

Information and software technology 2003.05 -Vol 45, No 7,  p. 419-29.

author: Chris Matthews

speaker: hsiu chun-chiu (9254610)

Page 2: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

2

Abstract(1/2)

Fuzzy logic toolkit – language ZProvides operators, measures and

modifier for the definition and manipulation of fuzzy sets and relations

A simple fuzzy expert system: use of fuzzy Cartesian product fuzzy set truncation operators offers a generic definition for a centroid

defuzzification function

Page 3: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

3

Z is a powerful analytical tool that facilitates system understanding through the development of a series of unambiguous, verifiable mathematical models

Z is a well-established specification languageThis example illustrates how the toolkit can

be used to specify a fuzzy rule base and provide the operators necessary for fuzzy inferencing.

Abstract(2/2)

Page 4: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

4

Fuzzy expert systems

rule based systemsincorporate fuzzy set theory and fuzzy

logic in both reasoning and knowledge representation

three major conceptual components A fuzzy rule base A description of the fuzzy set parameters A method for fuzzy inferencing (or reasoning)

Page 5: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

5

Mamdani Model

Linguistic rules: IF IF xx is is XXii and and yy is is YYii and … THEN and … THEN oo is is OOii

linguistic input variables : x , y

possible linguistic values : XXii , , YYii

linguistic output variables : oo

Page 6: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

6

For example (1/2)

A stock market advisory system

two input: share price and trading volume

linguistic variables

( stable, decreasing, heavy and moderate)

single output: decision( buy, sell or hold )

Page 7: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

7

example:IF share price is stable and trading volume is heavy THEN decision is hold

IF share price is decreasing and trading volume is moderate THEN decision is sell

For example (2/2)

Page 8: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

取mix

是利用股票價格及交易量來計算應該持有或銷售股票

Page 9: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

是利用股票價格及交易量來計算應該持有或銷售股票

Page 10: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

股票交易的決策圖

Page 11: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker
Page 12: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

12

Nine fuzzy rules & Cartesian products

Three of these lead to a decision to buy shares

Four to a decision to hold sharesTwo lead to a decision to sell sharesfuzzy Cartesian products( 笛卡兒乘積 )

Page 13: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

13

Rule base can be read as follows(1/2)

IF share price is stable and trading volume is light THEN decision is buy IF share price is increasing and trading volume is

moderate THEN decision is buy IF share price is increasing and trading volume is he

avy THEN decision is buy IF share price is stable and trading volume is heavy THEN decision is hold

Page 14: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

14

IF share price is stable and trading volume is moderate THEN decision is hold

IF share price is increasing and trading volume is light THEN decision is hold

IF share price is decreasing and trading volume is light THEN decision is hold

IF share price is decreasing and trading volume is

moderate THEN decision is sell

IF share price is decreasing and trading volume is heavy THEN decision is sell

Rule base can be read as follows(2/2)

Page 15: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

15

The initial system is

In the initial state all fuzzy sets and fuzzy relations, declared in the schema ruleBase are empty.

Page 16: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

16

BuildSharePriceis

input

Page 17: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

17

buildTradingVolume

written

Page 18: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

18

Advise(1/2)

Page 19: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

19

Advise(2/2)

Page 20: A formal specification for a fuzzy expert system 1 Information and software technology 2003.05 -Vol 45, No 7, p. 419-29. author : Chris Matthews speaker

A formal specification for a fuzzy expert system

20

Conclusion

This paper illustrates the use of the fuzzy logic toolkit in the specification of a simple fuzzy expert system

The toolkit definitions for fuzzy set union and the fuzzy Cartesian product operators are used to build these relations