You are on page 1of 4

International Journal for Basic Sciences and Social Sciences (IJBSS) ISSN: 2319-2968 Available Online at www.rfbss.org Vol.

1, Issue 3, November-2012 Page: 65-68

ON YAGERS PROBABILITY-POSSIBILITY TRANSFORMATIONS


Mamoni Dhar Department of Mathematics, Science College, Kokrajhar-783370, Assam, India E-mail: mamonidhar@rediffmail.com, mamonidhar@gmail.com
Abstract: In this article, our main intention is to discuss and comment on the existing probability-possibility consistency principles and thereby we would like to state that the works which are carried out with the help of these principles would yield unsatisfactory results. A new probability possibility consistency principle is suggested herein and it is expected that this will produce the required appropriate connection between probability and possibility. Keywords: Superimposition of Sets, Dubois and Prade Consistency Principles, Left Reference Function and Right reference function. 1. Introduction Possibility theory is a mathematical theory dealing with certain types of uncertainties and is considered as an alternative to probability theory. Possibility theory is devoted to the handling of incomplete information. The process of transformation from probability to possibility had received attention in the past. This question is philosophically interesting as a part of debate between probability and fuzzy sets. The conversion problem between probability and possibility has its roots in possibility probability consistency principle of Zadeh [14], that he introduced in the paper founding possibility theory. The transformation between probability and possibility has been studied by many researchers. Most of these studies examined principles that must be satisfied for transformations and derived an equation satisfying them. Dubios and Prade further contributed to its development. In Zadehs view, possibility distributions were meant to provide a graded semantics to natural statements. The transformation between probabilitiy to possibility is useful in some practical problems as: constructing a fuzzy membership function from a statistical data, ([12]), combining probabilities and possibilities in expert systems ([11]) and reducing complicated complexity ([9]). In other words, the transformation from possibility to probability or conversely is useful in case of decision making when the experts need precise information to take any decision. Yager [13] introduced a general procedure for transforming a probability distribution to another probabilistic distribution so that the resulting probability distribution at least has as much entropy as the original probability distribution. For developing this procedure, they used an approach of possibility probability distribution which was initially described by Dubois and Prade. The procedure for transformation from a probability distribution to another probabilistic distribution introduced by them can be described in the following manner: 1.1 Yagers general procedure for transformation Assume P is a probability distribution on X {x1, x2 ,.........x n } where p1 p2 ......... pn . The elements have been indexed in descending order of their probabilities. They then associated with these probability distributions a possibility distribution on X such that uj is the possibility of xj , where u n npn and u j j p j p j 1 u j 1 . It was mentioned that

using the formula stated above the trasformation from probability distribution to possibility distribution was derived. Similarly assuming u1 u 2 ......... u n as a normal probability distribution on X with u1=1, they obtained an associated probability distribution on X where
pn
u j u j 1 un and p j p j 1 j n

It is to be mentioned here that our intention is only to focus on the ideas underlying the procedures discussed but not on their technical details.

Mamoni Dhar, On Yagers Probability-Possibility Transformations

65

International Journal for Basic Sciences and Social Sciences (IJBSS) ISSN: 2319-2968 Available Online at www.rfbss.org Vol.1, Issue 3, November-2012 Page: 65-68
In this article, our main aim is to highlight that the above mentioned concepts would lead to some unsatisfactory results. The reason behind such a claim may be contributed to the fact that these results were based on Dubious and Prades conception of consistency between probability and possibility, which seems to have some shortcomings if it is seen from our perspective. They derived the consistency in the way because they were not satisfied with the way, Klir defined his consistency principle, but is it seen that there remained some questionable properties in Dubious and Prades consistency principles too. It is important to mention here that these well known and most commonly used consistency principles were derived on some misconceptions which can be explained in the form that possibility is described over the same space over which probability is described. It is important to mention here that Dubious and Prade defined a normal fuzzy number with the help of two functions of which one is on the left side of the point of maximum possibility and the other is on the right side of the point of maximum possibility (see for example [10]). Thus, it is seen that the membership function of a normal fuzzy number can be considered as a combination of two functions which they called left reference function and right reference function. But while defining a possible link between probability and possibility, they ignored their own definition of a normal fuzzy number by describing probability and possibility over the same space. These contradict the definition of normal fuzzy number proposed by the author himself. Again, we would like to note here that in the aforesaid transformations they used the term measure with possibility which is not at all appropriate in our view point. The reason for this can be explained as follows: The postulates that are required to define a measure are not followed in case of fuzzy sets. Further, it can be explained that in the way that the measure of a point is zero but the possibility is described by membership functions and hence it seems that the term measure with possibility do not sound good. Another thing to be mentioned here that the results were obtained only for discrete cases and nothing was mentioned for continuous cases. So the question arises, how much logical their principle is? At this juncture, we would like to suggest a new consistency principle which is expected to provide a very satisfactory result because the shortcomings that are observed in the existing principles will not be there in the proposed principle defined with the help of superimposition of sets. So while a number of measures have been suggested in the literature, we shall find the following one introduced in [1, 2] to be the most appropriate for our purpose. This principle was proposed by Baruah with the application of superimposition of sets and called it The Randomness- Fuzziness Consistency Principle. We prefer this principle because it is constructed on sound logic and appropriate mathematical frame works which is very much needed in anything dealing with mathematics. It is important to mention here that we have discussed about these in our previous works in [7, 8] and so we are not going to discuss in details about these in this work. But to make the matter clear and simple let us have a look at the principle in brief in the following section.

2. The randomness fuzziness consistency principles Baruah [2, 4] introduced a framework for reasoning with the link between probability possibility. The development of this principle focussed mainly on the existence of two laws of randomness which are required to define a law of fuzziness. In other words , not one but two laws of fuzziness is required to define a law of randomness on two disjoint spaces which in turn can construct a fuzzy membership function. Fundamental to this approach is the idea that possibility distribution can be viewed as a combination of distributions of which one is a probability distribution and the other is a complementary probability distribution. The consistency principle introduced in the manner can be explained mathematically in the following form: For a normal fuzzy number of the type N = [, , ] with membership function N(x) = 1(x), if x , = 2(x), if x , and = 0, otherwise, with 1 () = 2 () = 0, 1 () = 2 () = 1, the partial presence of a value x of the variable X in the interval [, ] is expressible as N(x) = Prob [ X x] + (1 ) {1 Prob [ X x]}, where = 1 if x and = 0 if x , (Baruah [3,5,6]). With the above result, we would like to discuss the fact that the spirit of this approach is to our opinion, better founded then the existing ones. If this be the case, then it is obvious that the results of all the transformations which basically depend on the existing link between probability and possibility or conversely would be illogical from our standpoints, the reasons for

Mamoni Dhar, On Yagers Probability-Possibility Transformations

66

International Journal for Basic Sciences and Social Sciences (IJBSS) ISSN: 2319-2968 Available Online at www.rfbss.org Vol.1, Issue 3, November-2012 Page: 65-68
which are discussed in the previous section. That is why the principles cannot be accepted for further studies and also those who depended on these results without having the in depth thinking would have to give a second thought to the procedures developed with the existing principles linking possibility with probability. In other words, we would like to say here that the method of linking probability with possibility which is suggested by us is preferable among various other existing transformation procedures because of its logical foundations and appropriate mathematical frameworks. While dealing with a subject like mathematics, it is very important to see whether the things which are in use are constructed within proper mathematical frameworks or not. It is necessary because otherwise we would have to be contended with some results having no logic at all. Hence it is expected that the above mentioned method of transformation would be workable in all respect and it is for this reason this principle of consistency is suggested in this article. 4. Conclusion In this article, we have explained a very important consistency principle to transform a probability distribution to possibility one or conversely. The main contribution of this article is to revisit some of the well- known consistency principles between probability and possibility. In the process, some of the principles are discussed and commented on and thereby we would like to say that the aforesaid principles are based on some misconceptions and hence all the results obtained on the basis of these principles have nothing to do. Henceforth, a new consistency principle is suggested because it is derived within an appropriate mathematical frameworks and it is expected to provide the most satisfactory results in the process of transformation between probability to possibility or conversely. It is important to note here that the principle introduced in the aforesaid manner is completely different from those found in the literature. If this be the case then it is obvious that the results that were obtained on the basis of existing consistency principles would lead to illogical results. That is to say that the proposed principle discard the process of finding a general procedure for transformation from a probability distribution to another possibilistic distribution which was introduced in a manner so that resulting probability distribution has as much entropy as the original probability distribution. The reason behind such a claim can be contributed to the fact that these were derived on the basis of Dubious and Prades conception of consistency between probability and possibility which itself does not obey the definition of normal fuzzy number defined by Dubois and Prade.

REFERENCES [1] Baruah H.K., Set Superimposition and Its Applications to The Theory of Fuzzy Sets, Journal of the Assam Science Society, Vol.40, No.1 & 2, pp.25-31, 1999. [2] Baruah H.K., The Randomness-Fuzziness Consistency Principles, IJEIC, Vol. 1, Issue 1, pp. 37-48, 2010. [3] Baruah H.K. , Construction of the Membership Function of a Fuzzy Number, ICIC Express Letters, Vol. 5, Issue 2, pp. 545-549, 2011. [4] Baruah H.K., Theory of Fuzzy sets: Beliefs and Realities, IJEIC, Vol. 2, Issue 2, pp.1-22, 2011. [5] Baruah H.K., In Search of the Root of Fuzziness: The Measure Theoretic Meaning of Partial Presence, Annals of Fuzzy Mathematics and Informatics, Vol. 2, No. 1, pp. 57 68, 2011. [6] Baruah H.K., Construction of Normal Fuzzy Numbers Using the Mathematics of Partial Presence, Journal of Modern Mathematics Frontiers, Vol.1, No. 1, pp. 9 - 15, 2012. [7] Dhar M., A Note on Coherence Between Probability and Possibility Measures, International Journal of Computer Applications, USA, 2012. [8] Dhar M., Probability-Possibility Transformations: A Brief Revisit, Annals of Fuzzy Mathematics and Informatics, Korea, 2012. [9] Dubois D. and Prade H., On possibility/probability transformations, Fuzzy Logic, pp. 103-112, 1993. [10] Kaufman A. and Gupta M.M., Introduction to Fuzzy Arithmetic, Theory and Applications, Van Nostrand Reinhold Co. Inc., Wokingham, Berkshire, 1984. [11] Klir G.J., Probability-Possibility Transformations : A Comparision, International Journal of General System Vol 21, pp. 291-310, 1992. [12] Krishnapuram R., Keller J.M., A Possibilistic Approach to Clustering, IEEE transaction on fuzzy systems, Vol.1, No.2, pp. 98-109, 1993.

Mamoni Dhar, On Yagers Probability-Possibility Transformations

67

International Journal for Basic Sciences and Social Sciences (IJBSS) ISSN: 2319-2968 Available Online at www.rfbss.org Vol.1, Issue 3, November-2012 Page: 65-68
[13] Yager R. R., Entropy Conserving Probability Transform and the entailment Principles, Fuzy Sets and Systems, Vol. 158, pp.1397-1405, 2007. [14] Zadeh L. A., Fuzzy Sets as a Basis For a Theory of Possibility, Fuzzy Sets and Systems, 1, pp.3-28, 1978.

Mamoni Dhar, On Yagers Probability-Possibility Transformations

68

You might also like