23.02.2015 Views

rst09 panel - Rough Sets Theory

rst09 panel - Rough Sets Theory

rst09 panel - Rough Sets Theory

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Semantics of <strong>Rough</strong> <strong>Sets</strong><br />

From theory to applications (through<br />

semantics understanding).<br />

Motivated by Dubois, D. and Prade, P.<br />

The three semantics of fuzzy set,<br />

Fuzzy <strong>Sets</strong> and Systems, 90, 141-150,<br />

1997.


<strong>Rough</strong> sets and concepts<br />

<strong>Rough</strong> sets for modeling concepts, or<br />

more precisely the extensions of<br />

concepts.


Concepts<br />

A concept, in the classical view, is<br />

defined by a pair of intension and<br />

extension.<br />

The extension: instances to which the<br />

concept applies.<br />

The intension: a set of singly necessary<br />

and jointly sufficient conditions that<br />

describe the instances of the concept.


Concepts: idealization<br />

Concepts are assumed to have welldefined<br />

boundaries and their<br />

extensions can be precisely defined by<br />

sets of objects.


Concepts: reality<br />

Concepts in many practical situations<br />

cannot be precisely defined and their<br />

extensions may not be given exactly by<br />

sets of objects.


Concepts: generalization<br />

Concepts with grey/gradual boundaries.<br />

Partially known concepts.<br />

Undefinable concepts.<br />

Approximate concepts.<br />

… … … …<br />

* Last three situations were discussed by Marek, V.W.<br />

and Truszczynski, M. Contributions to the theory of<br />

rough sets, Fundamenta Informaticae, 39(4) , 389 -<br />

409, 1999.


Concepts with grey/gradual boundaries<br />

<br />

There is not a well-defined boundary that<br />

differentiates the instances from the noninstances<br />

of the concept. For some objects,<br />

the concept only applies partially instead of<br />

fully.<br />

Fuzzy sets.<br />

* Zadeh, L.A. Fuzzy sets, Information and Control, 8, 338–353,<br />

1965.<br />

* Dubois, D. and Prade, P. Fuzzy rough sets and rough fuzzy<br />

sets, International Journal of General Systems, 17, 191-209.


Partially known concepts<br />

An object may actually be either an<br />

instance or not an instance of a<br />

concept.<br />

Due to a lack of information and<br />

knowledge, one can only express the<br />

state of instance and non-instance for<br />

some objects, instead of all objects.


Partially known concepts<br />

<br />

One has a partially known concept defined by a<br />

lower bound and upper bound of its extension.<br />

Interval sets<br />

<br />

Y.Y. Yao, Interval-set algebra for qualitative knowledge<br />

representation, Proceedings of the Fifth International<br />

Conference on Computing and Information, 370-374, 1993.


Undefinable concepts<br />

In general, the intension of a concept<br />

may be defined by using a logical<br />

language, such as the decision logic<br />

language in rough set theory and<br />

knowledge representation languages of<br />

description logic.


Undefinable concepts<br />

It may happen that some objects<br />

cannot be differentiated due to the use<br />

of a fixed and limited set of attributes<br />

for their description. The language<br />

may not be able to define certain sets<br />

of objects that are the extensions of<br />

some concepts. That is, we have<br />

some undefinable concepts with<br />

respect to a particular language..


Undefinable concepts<br />

<br />

We have to approximate undefinable<br />

concepts by definable concepts.<br />

<strong>Rough</strong> sets.<br />

<br />

Pawlak, Z. <strong>Rough</strong> sets, International Journal of<br />

Computer and Information Sciences, 11, 341-356,<br />

1982.


Approximate concepts<br />

One knows the exact extension of a<br />

concept and the language can<br />

precisely define the concept.<br />

The description may be too complex to<br />

be of any practical value; it may be<br />

difficult to understand and manipulate.


Approximate concepts<br />

<br />

This may require us to approximate the<br />

concept by some concepts with simple<br />

descriptions.<br />

<strong>Rough</strong> sets.<br />

* Marek, V.W. and Truszczynski, M. Contributions to<br />

the theory of rough sets, Fundamenta Informaticae,<br />

39(4) , 389 - 409, 1999.


Thank You!<br />

http://www.cs.uregina.ca/~yyao/rough_set.html

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!