Re: Consistency of fuzzy rules

From: WSiler@aol.com
Date: Wed Oct 31 2001 - 21:51:25 MET

  • Next message: WSiler@aol.com: "Re: Consistency of fuzzy rules"

    In a message dated 10/15/01 2:33:18 AM Central Daylight Time,
    ftrjin@mail.ipm.net writes:

    > Sure, fuzzy rules allows ambiguities, but fully contradictory rules not only
    > confuse users, but also degrade the performance of the fuzzy system.
    > Otherwise,
    > why do you need another rules to resolve the contradiction??? I don't think
    > a
    > fuzzy system will be easily accepted if the user find it is full of
    > contradictory fuzzy rules.

    I am afraid we have a difference of opinion. You are assume that your
    definition of contradictory rules is correct; I do not think this is
    necessarily the case. I do think that scanning the rules according to your
    definition and issuing a warning to the programmer of possibly contradictory
    rules is a good idea; however, I think that this case warrants an error
    message. For example, in a fuzzy control program, it is very common to have
    two perfectly correct rules such as

          If temperature is VeryHigh and fan is Medium then heat is Low
          If temperature is High and fan is High then heat is Low

    These rules are not contradictory; they simply say that there is more than
    one input state which can give rise to the same output state. Similarly, it
    is possible for the same input state to give rise to more than one output
    state. For example, the same set of symptoms may suggest more than one disease

    In simple fuzzy control, it is not uncommon to have only one block of rules
    to reach a decision. However, in general-purpose fuzzy reasoning, it is rare
    to have only one rule block which reaches a conclusion; in most cases,
    several (perhaps 10 or more rule blocks) passing through several reasoning
    steps before a conclusion is reached. For example, a system we have
    constructed for monitoring pulmonary artery cather pressures in a hospital
    intensive care unit have 14 reasoning steps, each a block of several rules. A
    number of these rules would not pass your tests for consistency.
     
    Sincerely, William Siler

    --part1_103.ae4936b.29057d5b_boundary
    Content-Type: text/html; charset="US-ASCII"
    Content-Transfer-Encoding: 7bit

    <HTML><FONT FACE=arial,helvetica><FONT SIZE=2>In a message dated 10/15/01 2:33:18 AM Central Daylight Time, ftrjin@mail.ipm.net writes:
    <BR>
    <BR>
    <BR><BLOCKQUOTE TYPE=CITE style="BORDER-LEFT: #0000ff 2px solid; MARGIN-LEFT: 5px; MARGIN-RIGHT: 0px; PADDING-LEFT: 5px">Sure, fuzzy rules allows ambiguities, but fully contradictory rules not only
    <BR>confuse users, but also degrade the performance of the fuzzy system. Otherwise,
    <BR>why do you need another rules to resolve the contradiction??? I don't think a
    <BR>fuzzy system will be easily accepted if the user find it is full of contradictory fuzzy rules. </FONT><FONT COLOR="#000000" SIZE=3 FAMILY="SANSSERIF" FACE="Arial" LANG="0"></BLOCKQUOTE>
    <BR></FONT><FONT COLOR="#000000" SIZE=2 FAMILY="SANSSERIF" FACE="Arial" LANG="0">
    <BR>I am afraid we have a difference of opinion. You are assume that your definition of contradictory rules is correct; I do not think this is necessarily the case. I do think that scanning the rules according to your definition and issuing a <B>warning </B>to the programmer of <B>possibly</B> contradictory rules is a good idea; however, I think that this case warrants an <B>error </B>message. For example, in a fuzzy control program, it is very common to have two perfectly correct rules such as
    <BR>
    <BR> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;If temperature is VeryHigh and fan is Medium then heat is Low
    <BR> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;If temperature is High and fan is High then heat is Low
    <BR>
    <BR>These rules are not contradictory; they simply say that there is more than one input state which can give rise to the same output state. Similarly, it is possible for the same input state to give rise to more than one output state. For example, the same set of symptoms may suggest more than one disease
    <BR>
    <BR>In simple fuzzy control, it is not uncommon to have only one block of rules to reach a decision. However, in general-purpose fuzzy reasoning, it is rare to have only one rule block which reaches a conclusion; in most cases, several (perhaps 10 or more rule blocks) passing through several reasoning steps before a conclusion is reached. For example, a system we have constructed for monitoring pulmonary artery cather pressures in a hospital intensive care unit have 14 reasoning steps, each a block of several rules. A number of these rules would not pass your tests for consistency.
    <BR>
    <BR>Sincerely, William Siler</FONT></HTML>

    --part1_103.ae4936b.29057d5b_boundary--

    ############################################################################
    This message was posted through the fuzzy mailing list.
    (1) To subscribe to this mailing list, send a message body of
    "SUB FUZZY-MAIL myFirstName mySurname" to listproc@dbai.tuwien.ac.at
    (2) To unsubscribe from this mailing list, send a message body of
    "UNSUB FUZZY-MAIL" or "UNSUB FUZZY-MAIL yoursubscription@email.address.com"
    to listproc@dbai.tuwien.ac.at
    (3) To reach the human who maintains the list, send mail to
    fuzzy-owner@dbai.tuwien.ac.at
    (4) WWW access and other information on Fuzzy Sets and Logic see
    http://www.dbai.tuwien.ac.at/ftp/mlowner/fuzzy-mail.info
    (5) WWW archive: http://www.dbai.tuwien.ac.at/marchives/fuzzy-mail/index.html



    This archive was generated by hypermail 2b30 : Wed Oct 31 2001 - 22:07:53 MET