Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Available online at www.ispacs.com/jfsva Volume 2012, Year 2012 Article ID jfsva-00051, 9 pages doi:10.5899/2012/jfsva-00051 Research Article The Median Value of Fuzzy Numbers and its Applications in Decision Making Rahim Saneifard 1∗ , Rasoul Saneifard2 (1) Department of Applied Mathematics, Urmia Branch, Islamic Azad University, Oroumieh, Iran. (2) Department of Engineering Technology, Texas Southern University, Houston, Texas, USA. c Rahim Saneifard and Rasoul Saneifard. This is an open access article distributed under Copyright 2011 ⃝ the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract In this paper, the researchers proposed a new method to rank fuzzy numbers by median value of fuzzy numbers. The median value can be used as a crisp approximation with respect to a fuzzy quantity. Therefore, a method for ordering the fuzzy numbers is defined accordingly. This method can effectively rank various fuzzy numbers, their images and overcome the shortcomings of the previous techniques. Keywords : Ranking; Fuzzy numbers; Defuzzification; Interval-Value; Central interval. 1 Introduction In many applications, ranking of fuzzy numbers is an important component of the decision process. In addition to a fuzzy environment, ranking is a very important decision making procedure. Since Jain [9, 10] employed the concept of maximizing set to order the fuzzy numbers in 1976 (1978), many authors have investigated various ranking methods. Some of these ranking methods have been compared and reviewed by Bortolan and Degani [2], and more recently by Chen and Hwang [4]. Other contributions in this field include: an index for ordering fuzzy numbers defined by Choobineh and Li [6], ranking alternatives using fuzzy numbers studied by Dias [7], automatic ranking of fuzzy numbers using artificial neural networks proposed by Requena et al. [11], ranking fuzzy values with satisfaction function investigated by Lee et al. [5], ranking and defuzzification methods based on area compensation presented by Fortemps and Roubens [8], and ranking alternatives with fuzzy weights using maximizing set and minimizing set given by Raj and ∗ Corresponding author. Email address: [email protected] Tel:+989149737077 1 2 Journal of Fuzzy Set Valued Analysis Kumar [12]. However, some of these methods are computationally complex and difficult to implement, and others are counterintuitive and not discriminating. Furthermore, many of them produce different ranking outcomes for the same problem. In 1988, Lee and Li [13], proposed a comparison of fuzzy numbers by considering the mean and dispersion (standard deviation) based on the uniform and the proportional probability distributions. Having reviewed the previous methods, this article proposes a method to use the concept of median value, so as to find the order of fuzzy numbers. This method can distinguish the alternatives clearly. The main purpose of this article is that, the median value can be used as a crisp approximation of a fuzzy number. Therefore, by the means of this defuzzification, this article aims to present a new method for ranking of fuzzy numbers. In addition to its ranking features, this method removes the ambiguities resulted from the comparison of previous ranking. In Section 2, we recall some fundamental results on fuzzy numbers. In Section 3, a crisp approximation of a fuzzy number is obtained. In this Section some theorems and remarks are proposed and illustrated. Proposed method for ranking fuzzy numbers is in the Section 4. 2 Basic Definitions and Notations The basic definitions of a fuzzy number are given in [13, 14, 15, 16, 17, 18, 19, 20, 21, 22] as follows: Definition 2.1. A fuzzy number A is a mapping µA (x) : ℜ→ [0, 1] with the following properties: 1. µA is an upper semi-continuous function on ℜ, 2. µA (x) = 0 outside of some interval [a1 , b2 ] ⊂ ℜ. 3. There are real numbers a2 , b1 such that a1 ≤ a2 ≤ b1 ≤ b2 and 3.1 µA (x) is a monotonic increasing function on [a1 , a2 ], 3.2 µA (x) is a monotonic decreasing function on [b1 , b2 ], 3.3 µA (x) = 1 for all x in [a2 , b1 ]. The set of all fuzzy numbers is denoted by F . Definition 2.2. Let ℜ be the set of all real numbers. We assume a fuzzy number A that can be expressed for all x ∈ ℜ in the form ( )n x−a g(x) = when x ∈ [a, b), b−a 1, when x ∈ [b, c], )n ( A(x) = (2.1) d−x , when x ∈ (c, d], h(x) = d−c 0 otherwise . where a, b, c, d are real numbers such that a < b ≤ c < d and g is a real valued function that is increasing and right continuous and h is a real valued function that is decreasing and left continuous. A fuzzy number A with shape function g and h, where n > 0, will be denoted by A = ⟨a, b, c, d⟩n . If n = 1, we simply write A = ⟨a, b, c, d⟩, which is known as 3 Journal of Fuzzy Set Valued Analysis a trapezoidal fuzzy number. If n ̸= 1, a fuzzy number A∗ = ⟨a, b, c, d⟩n is a concentration of A. If 0 < n < 1, then A∗ is a dilation of A. Concentration of A by n = 2 is often interpreted as the linguistic hedge ”very”. Dilation of A by n = 0.5 is often interpreted as the linguistic hedge ”more or less”. More about linguistic hedges can be found in [22]. Another important notion connected with fuzzy number A is an cardinality of a fuzzy number A. Definition 2.3. [1]. Cardinality of a fuzzy number A described by (2.1) is the value of the integral ∫ b ∫ 1 cardA = A(x)dx = (bα − aα )dα. (2.2) a 0 If A = ⟨a, b, c, d⟩n then b−a d−c + (c − d) + . n+1 n+1 In this paper we will always refer to fuzzy number A described by (2.1). cardA = 3 (2.3) Median value The median value of a fuzzy set A was introduced as a possible scalar representative value of A. The researcher will adopt the description of the median value from [20]. Definition 3.1. [1]. The median value of a fuzzy number A characterized by (2.1) is the real number mA from the support of A such that ∫ mA ∫ d A(x)dx, (3.4) A(x)dx = a mA For practical purposes expression (3.4) can be rewritten as ∫ mA A(x)dx = 0.5 cardA. (3.5) a The article can classify fuzzy numbers with respect to the ”distribution” of their cardinality as follows: a fuzzy number A is called ∫b ∫d 1. a fuzzy number with equally heavy tails if a A(x)dx = c A(x)dx, } {∫ ∫d ∫d b 2. a fuzzy number with light tails if max a A(x)dx, c A(x)dx ≤ 0.5 a A(x)dx, 3. a fuzzy number with heavy left tail if ∫b 4. a fuzzy number with heavy right tail if a A(x)dx > 0.5 ∫d c ∫d A(x)dx > 0.5 a A(x)dx, ∫d a A(x)dx. Now the article will study location of the median value mA in the support of A. The article will also identify the fuzziness of mA determined by its membership grade A(mA ). Proposition 3.1. [1]. If A is a fuzzy number with light tails then (∫ d ) ∫ b b+c mA = + 0.5 A(x)dx − A(x)dx , 2 c a and A(mA ) = 1. (3.6) 4 Journal of Fuzzy Set Valued Analysis Corollary 3.1. If A has equally heavy tails then mA = b+c 2 and A(mA ) = 1. Obviously, if A has a heavy left (right) tail then mA is not located in the core of A but it is shifted to the left (right) end of the support of A. For a trapezoidal fuzzy number A and for its modifications by selected linguistic hedges the article will provide formulas for the location of the median value. Proposition 3.2. [1]. Let A = ⟨a, b, c, d⟩n . Then ( mA = a + (b − a)n (n + 1) cardA 2 ) 1 (n+1) , (3.7) , (3.8) if A has a heavy left tail, and ( mA = d − (d − c)n (n + 1) cardA 2 ) 1 (n+1) if A has a heavy right tail. Corollary 3.2. Let A = ⟨a, b, c, d⟩ be a trapezoidal fuzzy number. Then mA = a + √ (b − a) cardA, (3.9) (d − c) cardA, (3.10) if A has a heavy left tail, and mA = d − √ if A has a heavy right tail. When the median value is located outside the core of a fuzzy number A, it belongs to A with the membership grade less than 1 [1]. 4 Comparison of fuzzy numbers using a median value In the following, we present a new approach for ranking fuzzy numbers based on the distance method. The method not only considers the median value of a fuzzy number, but also considers the minimum crisp value of fuzzy numbers. For ranking fuzzy numbers, this study firstly defines a minimum crisp value τmin to be the benchmark and its characteristic function µτmin (x) is as follows: { 1, when x = τmin , µτmin (x) = (4.11) 0, when x ̸= τmin . When ranking n fuzzy numbers A1 , A2 , · · · , An , the minimum crisp value τmin is defined as: τmin = min{x|x ∈ Domain(A1 , A2 , · · · , An )}. (4.12) Assume that there are n fuzzy numbers A1 , A2 , · · · , An . The proposed method for ranking fuzzy numbers A1 , A2 , · · · , An is now presented as follows: Use the point (mAj , 0) to calculate the ranking value MAj = dist(mAj , τmin ) of the fuzzy numbers Aj , where Aj , where 1 ≤ j ≤ n, as follows: dist(mAj , τmin ) = ∥mAj − τmin ∥ (4.13) Journal of Fuzzy Set Valued Analysis 5 From formula (4.13), we can see that MAj = dist(mAj , τmin ) can be considered as the Euclidean distance between the point (mAj , 0) and the point (τmin , 0). We can see that the larger the value of MAj , the better the ranking of Aj , where 1 ≤ j ≤ n. Definition 4.1. Let, there are n fuzzy numbers A1 , A2 , · · · , An ., and MAj are the median value of their. Define the ranking of Aj by M(.) on F , i.e. 1. MAi < MAj if onlyif Ai ≺ Aj , 2. MAi > MAj if onlyif Ai ≻ Aj , 3. MAi = MAj if onlyif Ai ∼ Aj . Then, this article formulates the order ≽ and ≽ as Ai ≽ Aj if and only if Ai ≻ Aj or Ai ∼ Aj , Ai ≼ Aj if and only if Ai ≺ Aj or Ai ∼ Aj . 4.1 Using The Proposed Ranking Method In Selecting Army Equip System From experimental results, the proposed method with some advantages: (a) without normalizing process, (b) fit all kind of ranking fuzzy number, (c) correct Kerre’s concept. Therefore we can apply median value of fuzzy ranking method in practical examples. In the following, the algorithm of selecting equip systems is proposed, and then adopted to ranking a army example. 4.1.1 An algorithm for selecting equip system We summarize the algorithm for evaluating equip system as below: Step 1: Construct a hierarchical structure model for equip system. Step 2: Build a fuzzy performance matrix Ã. We compute the performance score of the sub factor, which is represented by triangular fuzzy numbers based on expert’s ratings, average all the scores corresponding to its criteria. Then, build a fuzzy performance matrix Ã. Step 3: Build a fuzzy weighting matrix W̃ . According to the attributes of the equip systems, experts give the weight for each criterion by fuzzy numbers, and then form a fuzzy weighting matrix W̃ . Step 4: Aggregate evaluation. To multiple fuzzy performance matrix and fuzzy weighting matrix W̃ , then get fuzzy aggregative evaluation matrix R̃. (i.e. R̃ = à ⊗ W̃ t ). Step 5: Determinate the best alternative. After step 4, we can get the fuzzy aggregative performance for each alternative, and then rank fuzzy numbers by median value of fuzzy numbers. 6 Journal of Fuzzy Set Valued Analysis Table 1 Linguistic values for the ratings. Linguistic value Very Poor(VP) Poor Slightly(SP) Fair(F) Slightly good(SG) Good(G) Very good(VG) TFNs (0,0,0.16) (0,0.16,0.33) (0.16,0.33,0.5) (0.33,0.5,0.66) (0.5,0.66.0.83) (0.66,0.83,1) (0.83,1,1) Table 2 Basic performance data for five types of main battle Tanks. Item Type Tank A Armament Ammunition 120 mm gun 15.2 mm MG 12.7 mm MG 40 1000 Tank B Tank C Tank D Tank E 120 mm gun 15.2 mm MG 120 mm gun 15.2 mm MG 105 mm gun 15.2 mm MG 120mm gun 7.62mm MG 12.7 mm MG Up to 50 4000 42 40 44 4750 4 1500 10000 2×6 2×5 2×8 None 2×9 26.2 19.2 27.2 19.0 27.5 67 56 72 60 71 480 1.21 60 2.74 Good Good Fair Medium 450 1.07 60 2.43 Excellent Fair Fair Medium 550 1.0 60 3.00 Good Good Fair Medium 300 1.2 55 2.51 Fair Fair Poor Medium 550 1.23 60 2.92 Excellent Good Fair Good 11400 Smoke grenade discharges Power to weight ratio(hp/t) Max. road speed(km/h) Max. range(km) Fording(m) Gradient Trench Armor protection Acclimatization Communication Scout 4.1.2 The selecting of best main battle tank In [3], the authors have constructed a practical example for evaluating the best main battle tank, and they selected x1 = M1 A1 (USA), x2 = Challenger2 (UK), x3 = Leopard2 (Germany) as alternatives. In [3], the experts opinion were described by linguistic terms, which can be repressed in triangular fuzzy numbers. The fuzzy Delphi method is adopted to adjust the fuzzy rating of each expert to achieve the consensus condition. The evaluating criteria of main battle tank are a1 : attackcapability, a2 : mobilitycapability, a3 : self − def encecapability and a4 : communicationandcontrolcapability. In this example, we adopted the hierarchical structure constructed in [3] for selection of five main battle tanks, and the step-by-step illustrations based on Sec. 4.1.1s algorithm are described bellow: 7 Journal of Fuzzy Set Valued Analysis Step 1: Construct a hierarchical structure model for equip system. Step 2: Build a fuzzy performance matrix Ã. The basic performance data for five types of main battle tanks are summarized in Table 2. Then based on the linguistic values in Table 1, the fuzzy preference of five tanks toward four criteria are collected and shown in Table 5. Step 3: Build a fuzzy weighting matrix W̃ . The aggregative fuzzy weights of four criteria, according to the linguistic values of importance in Table 3, are shown in Table 4. Step 4: Aggregate evaluation. To multiple fuzzy performance matrix à and fuzzy weighting matrix W̃ , then get fuzzy aggregative evaluation matrix R̃ = à ⊗ W̃ t . Therefore, from Table 4 and 5, we have R̃ = (0.7, 0.9, 1.0) (0.7, 0.8, 1.0) (0.4, 0.6, 0.7) (0.4, 0.6, 0.8) (0.6, 0.8, 0.9) (0.7, 0.9, 0.96) (0.2, 0.3, 0.5) (0.3, 0.5, 0.6) (0.7, 0.8, 0.9) (0.8, 0.9, 1.0) = (0.5, 0.7, 0.8) (0.5, 0.7, 0.8) (0.4, 0.6, 0.8) (0.3, 0.5, 0.7) (0.6, 0.8, 0.9) (0.6, 0.8, 0.9) (0.6, 0.8, 0.9) (0.6, 0.8, 0.9) (0.4, 0.6, 0.8) (0.7, 0.8, 1.0) (0.38, 0.61, 0.82) (0.27, 0.48, 0.69) (0.36, 0.58, 0.77) (0.18, 0.34, 0.55) (0.40, 0.64, 0.84) (0.7, 0.9, 1.0) (0.8, 0.9, 1.0) ⊗ (0.6, 0.7, 0.8) (0.3, 0.5, 0.7) . Step 5: Determinate the best alternative. According to Eq. (4.13), we can get the median value of fuzzy numbers of Tanks A-E, which are equal to 0.234, 0.423, 0.236, 0.323and0.289, respectively. Therefore, we find that the ordering of median value is T ank A < T ank C < T ank F < T ank D < T ank B. So, the best type of main battle Tank is Tank F . Table 3 Linguistic values of the importance weights. Linguistic value TFNs Very low(VL) (0,0,0.167) Low(L) (0,0.167,0.333) Slightly (0.167,0.333,0.5) Medium(M) (0.333,0.5,0.667) Slightly high(SH) (0.5,0.667,0.833) High(H) (0.667,0.833,1) Very High(VH) (0.833,1,1) Table 4 The importance weights of linguistic criteria and Criteria Experts D1 Attack (W̃1 ) VH Mobility (W̃2 ) VH Self-defence (W̃3 ) M Communication and command (W̃4 ) M its mean. Mean of TFNs D2 H H VH M D3 H VH SH M (0.72,0.89,1) (0.78,0.94,1) (0.56,0.72,0.83) (0.33,0.5,0.67) 8 Journal of Fuzzy Set Valued Analysis Table 5 Basic performance data for five types of main battle Tanks. Criteria Type Tank A Tank B Tank C Attack Armament G SG SG Ammunition VG SG SG Smoke grenade G SP VG dischargers Mean (0.7,0.8,1) (0.3,0.5,0.7) (0.6,0.7,0.8) Mobility Power to weight G F G ratio Max. G F VG road speed Max. range G SG VG Fording/Gradient G SG SG Trench Mean (0.6,0.8,1) (0.4,0.5,0.7) (0.7,0.8,0.9) Self-defence Armor SG G F protection Acclimatization SG F SG Mean (0.5,0.6,0.8) (0.5,0.6,0.8) (0.4,0.5,0.7) Communication and command Communication G G G Scout SG SG SG Mean (0.5,0.7,0.9) (0.5,0.7,0.9) (0.5,0.7,0.9) 5 Tank D Tank E F F VP SG G VG (0.2,0.3,0.5) (0.6,0.8,0.9) F G SG VG P F VG G (0.2,0.4,0.6) (0.7,0.9,1) F G F (0.3,0.5,0.6) G (0.5,0.7,0.9) F SG (0.4,0.5,0.7) G G (0.6,0.8,1) Conclusion In this paper, the researchers proposed a median value method to rank fuzzy numbers. This method use a crisp set approximation of a fuzzy number. The method can effectively rank various fuzzy numbers and their images (normal/nonnormal/trapezoidal/general ). The calculations of this method are simpler than the other approaches. References [1] S. Bodjanova, Median value and median interval of a fuzzy number, Infomation Sciences 172 (2005) 73-89. http://dx.doi.org/10.1016/j.ins.2004.07.018. [2] G. Bortolan and R. Degani, A review of some methods for ranking fuzzy numbers, Fuzzy Sets and Systems 15 (1985) 1-19. http://dx.doi.org/10.1016/0165-0114(85)90012-0. [3] CH. Cheng, Y. Lin, Evaluating weapon system by analitic hierarchy process based on fuzzy scales, Fuzzy Sets and Systems 63 (2002) 1-10. [4] S. J. Chen and C. L. Hwang, Fuzzy Multiple Attribute Decision Making, Springer- Verlag, Berlin, 1972. Journal of Fuzzy Set Valued Analysis 9 [5] K. Lee, H. Chu, H. Lee-Kwang, Ranking fuzzy values with satisfaction function, Fuzzy Sets and Systems 64 (1994) 295-309. http://dx.doi.org/10.1016/0165-0114(94)90153-8. [6] F. Choobineh, H. Li, An index for ordering fuzzy numbers, Fuzzy Sets and Systems 54 (1993) 287-294. http://dx.doi.org/10.1016/0165-0114(93)90374-Q. [7] O. Dias, Ranking alternatives using fuzzy numbers: A computational approach, Fuzzy Sets and Systems 56 (1993) 247-252. http://dx.doi.org/10.1016/0165-0114(93)90151-7. [8] P. Fortemps, M. Roubens, Ranking and defuzzification methods based on area compensation, Fuzzy Sets and Systems 82 (1996) 319-330. http://dx.doi.org/10.1016/0165-0114(95)00273-1. [9] R. Jain, Decision-making in the presence of fuzzy variable, IEEE Trans. Systems Man and Cybernet.SMC. 6 (1976) 698-703. http://dx.doi.org/10.1109/TSMC.1976.4309421. [10] R. Jain, A procedure for multi-aspect decision making using fuzzy sets, Internat. J. Systems Sci. 8 (1978) 1-7. http://dx.doi.org/10.1080/00207727708942017. [11] I. Requena, M. Delgado and J. I. Verdegay, Automatic ranking of fuzzy numbers with the criterion of decision maker leant by an artificial neural network, Fuzzy Sets and Systems 64 (1994) 1-9. http://dx.doi.org/10.1016/0165-0114(94)90002-7. [12] P. A. Raj, D. N. Kumar, Ranking alternatives with fuzzy weights using maximizing set and minimizing set, Fuzzy Sets and Systems 105 (1999) 365-375. http://dx.doi.org/10.1016/S0165-0114(97)00243-1. [13] R. Saneifard, Defuzzification method for solving fuzzy linear systems, International Journal of Industrial Mathematics 4 (2009) 321-331. [14] R. Saneifard, Ranking L-R fuzzy numbers with weighted averaging based on levels, International Journal of Industrial Mathematics 2 (2009) 163-173. [15] R. Saneifard, T. Allahviranloo, F. Hosseinzadeh, N. Mikaeilvand, Euclidean ranking DMUs with fuzzy data in DEA, Applied Mathematical Sciences 6 (2007) 2989-2998. [16] R. Saneifard, A method for defuzzification by weighted distance, International Journal of Industrial Mathematics 3 (2009) 209-217. [17] R. Saneifard, R. Ezatti, Defuzzification Through a Bi Symmetrical weighted Function, Aust. J. Basic Appl. Sci. 10 (2010) 4976-4984. [18] Rahim Saneifard and Rasoul Saneifard, A modified method for defuzzification by probability density function, Journal of Applied Sciences Research 2 (2011) 102-110. [19] R. Saneifard and A. Asgari, A Method For Defuzzification Based on Probability Density Function (II), Applied Mathematical Sciences 5 (2011) 1357-1365. [20] Rahim Saneifard and Rasoul Saneifard, Evaluation of Fuzzy Linear Regression Models by Parametric Distance, Australian Journal of Basic and Applied Sciences 5 (2011) 261-267. [21] L. A. Zadeh, Fuzzy sets, Inform. Control 8 (1965) 338-353. http://dx.doi.org/10.1016/S0019-9958(65)90241-X. [22] L. A. Zadeh, A fuzzy-set-theoretic interpretation of linguistic hedges, Journal of Cybernetics 2 (1972) 4-34. http://dx.doi.org/10.1080/01969727208542910.