Fuzzy Relations “My crystal ball is fuzzy.” ~ Lotfi A. ZadehIntroduction @Relation: OHas the computational potency and significance in fuzzy approaches that functions have in conventional approaches Olmplies the presence of an association between elements of different sets * Crisp Relations: The degree of association is either 0 or 1 + Fuzzy Relations: The degree of association is a number between 0 and 1Cartesian Product of Relation @The Cartesian product of two sets X (for example the points on an x-axis) and Y (for example the points on a y-axis), denoted X x Y, is the set of all possible ordered pairs whose first component is a member of X and whose second component is a member of Y. For example: @A = {0,1}; B = {a, b, c} XxY={(x,y)|ceX andyeY}. A x B= {(0,a), 0, b), (0,c), (1a), (1,5), (1, 0} B x A= {(a,0), (a, 1), (b, 0), (b, 1), (ce, 0), (e, D}. AxA= {(0, 0), (0, 1), 1,0), 1, D} B x B = B? = {(a, a), (a,b), (a,c), (ba), (bb), (b, ©), (€. a), (Cb), (C0)Crisp Relations @A relation among classical sets x;, Xz, ..., X, aNd Vy, Yo, ---, Yn iS a subset of the Cartesian product. @\t is denoted either by R or by the abbreviated form X x Y = {(x, y)/x € X, y € Y} which forms an ordered pair of every x € X with every y € Y, forming unconstrained matches between X and Y. @That is, every element in universe X is related completely to every element in universe Y.Crisp Relations @In the case of an ordered pair, the relation is a subset of the Cartesian product A, x Ap. @This subset of the full Cartesian product is called as the binary relation from A, into A>.Crisp Relations @The strength of the relationship between ordered pairs of elements in each universe is measured by the characteristic function denoted by X, where a value of unity is associated with complete relationship and a value of zero is associated with no relationship, ie., 1 (z,y)¢€XxY, Xexy (4) = ie (x,y) EX xY.Crisp Relations @When the universe or the set are finite, a matrix called as relation matrix can conveniently represent the relation. A two-dimensional matrix represents the binary relation. elf X = {2, 4, 6} and Y = {p, q, #}, if they both are related to each other entirely, then the relation between them can be given by: Pp rerrererar) r I 1 1 x Il orp 1 1 1Crisp Relations @A more general crisp relation, R, exists when matches between elements in two universes are constrained. @Again, the characteristic function is used to assign values of relationship in the mapping of the Cartesian space X x Y to the binary values of (0, 1): 1, (@,yeER ARG Y) = {0 (x,y) ¢R’Crisp Relations @Example: @Let R be a relation among the two sets X = {1, 2} and for Y = {a, b} R= {(1, a), (2, b)} @Solution: The relation can be represented as follows: abUnion Intersection Complement Containment Identity Operations on Crisp Relations RUS — pauslt,y)< anus y) = marl yar, y), xslt, RAS — parstt,y)< xaos y) = mye, y), x90] R— gly) play) == pg) RCS jest): RC) $ 4908,)) 00a XE.Properties of Crisp Relations @The properties of commutativity, associativity, distributivity, involution, and idempotency for the classical sets also hold good for crisp relation. @This includes DeMorgan’s laws and the excluded middle laws too. The null relation O and the complete relation E are given by: 0 0 0 1 O=|0 0 0], E=]1 00 0 1Composition @Let R be relation that relates elements from universe X to universe Ys @Let S be the relation that relates elements from universe Y to universe Z. @Let T relates the same element in universe that R contains to the same elements in the universe Z that S contains. @The two methods of the composition operations are: —Max-—min composition, — Max-product composition.Composition @The max-min composition is defined by the set-theoretic and membership function-theoretic expressions: T= mp(x,y) S Hs(x,y).Properties of Fuzzy Relations eJust as for crisp relations, the properties of commutativity, associativity, distributivity, involution, and idempotency all hold for fuzzy relations. ®@Moreover, De Morgan’s principles hold for fuzzy relations just as they do for crisp (classical) relations, and the null relation, O, and the complete relation, E, are analogous to the null set and the whole set in set-theoretic form, respectively.Properties of Fuzzy Relations @As in case of fuzzy set, in fuzzy relation also excluded middle law and contradiction law does not holds good. @Since a fuzzy relation ? is also a fuzzy set, there is overlap between a relation and its complement; hence, RUR#E. RORZO.Fuzzy Relations (I) @There is an ambiguity in degree of association of elements @Fuzzy relations are fuzzy sets defined on Cartesian product OCartesian Product: Simply a high dimensional universe of discourse + Example: Xx Y, Xx Y x Z, ete. OFuzzy sets: defined on a single universe of discourse (e.g., X) OFuzzy relations: defined on higher-dimension universe of discourses (e.g., Xx X, Xx Y, etc.)Fuzzy Relations (Il) @Fuzzy Cartesian Product: DA relation between two or more fuzzy sets OExample: A: a fuzzy set on universe X B: a fuzzy set on universe Y Cartesian product between A and B: a fuzzy relation AxB = Rc X x¥ where: MOV) = Mace Y) = min( {4 (2), Hy (V))Fuzzy Relations (Ill) @Example: Fuzzy set A defined on a universe of three discrete temperatures, X= (X,, X., Xs} aS: 02 05,10 4 = ec) Fuzzy set B defined on a universe of two discrete pressures, Y={y,, ¥z} as: 03 09 Cartesian product AxB: Ye Ve »f02 0.2 x,|0.3 09Composition of Fuzzy Relations Fuzzy composition can be defined just as it is for crisp (binary) relations. @Suppose # is a fuzzy relation on the Cartesian space X x Y, S$ is a fuzzy relation on Y x Z, and 7 is a fuzzy relation on X x Z, then fuzzy max-min composition is defined as: T = R 0S (set-theoretic notation) = {(e.2),myx {nin (unload sas (y, )}/s eXyey v : 3 @\n membership function-theoretic notation er (%,2) = V (ue (x, y)A Hs (y, yéComposition of Fuzzy Relations @\n fuzzy max—product composition is defined in terms of set-theoretic T=RoS = {ema (ia(0.0 +1502) [es XyeY,z¢ a\ v \ha : In membership function-theoretic notation ur (z,z)= V (He (e0)e1s (n= ) yey ~ ~Composition of Fuzzy Relations @ Example: OGiven X = {x1, x2}, Y= {y1, y2}, and Z = (z1, z2, z3} Consider the following fuzzy relations: vt 2 21 22 z x; [0.7 05 ~ 9 190.9 06 0.2 ” Ee Al and $=), et 07 a Let T relates elements of universe X to elements of universe Z, that is, defined on Cartesian space X x Z @max—min composition example eT, 21) = max[min(0.7, 0.9), min(0.5, 0.1)] = 0.7 @Try to complete for T R=Composition of Fuzzy Relations ‘Example: UGiven x = {x1, x2}, Y= {y1, y2}, and Z = (21, 22, 23} Consider the following fuzzy relations: M 2 Z 22 z3 _—« [07 05) | 1 [0.9 06 0.2 R= Se ie o4| 4 S=> lon 07 05 Let T relates elements of universe X to elements of universe Z, that is, defined on Cartesian space X x Z @max-product composition example He7(x2, 22) = max[ (0.8 - 0.6), (0.4 -0.7)] = 0.48 @Try to complete for T.Tolerance and Equivalence Relations @Relations can exhibit various useful properties Relations can be used in graph theory @A relation R on a universe X can also be thought of as a relation from X to X. The relation R is an equivalence relation if it has the following three properties: (1) reflexivity, (2) symmetry, and (3) transitivity. a2 Three-vertex graphs for properties of (a) reflexivity, (b) symmetry, and (c) transitivityCrisp Equivalence Relations @Reflexivity — every vertex in the graph originates a single loop @For matrix relation reflexivity is given by (a;,a;)€R or Xp(x;,a;) =1.Crisp Equivalence Relations Symmetry - in the graph for every edge pointing (the arrows on the edge lines) from vertex j to vertex j (i, j = 1, 2, 3), there is an edge pointing in the opposite direction, that is, from vertex j to vertex i . @For matrix relation symmetry is given by (wi,2;) € R= (@j,a;) ER or xn(wi,25) = xn(2j,21)-Crisp Equivalence Relations Transitivity - for every pair of edges in the graph, one pointing from vertex i to vertex j and the other from vertex j to vertex k (i, j, kK = 1, 2, 3), there is an edge pointing from vertex / directly to vertex k For matrix relation transitivity is given by (wi,0j;)€R and (a;,2x) € R— (ai,2%) ¢R XR(ti,2j) and XA(xj,ex) =1—> VR(%, ex) = 1.Crisp Tolerance Relation @A tolerance relation R (also called a proximity relation) on a universe X is a relation that exhibits only the properties of reflexivity and symmetry. @A tolerance relation, R, can be reformed into an equivalence relation by at most (n - 1) compositions with itself, where n is the cardinal number of the set defining R, in this case X, that is, Rt = Ro Ryo +. oR =R.Crisp Tolerance Relation Suppose in an airline transportation system we have a universe composed of five elements: the cities Omaha, Chicago, Rome, London, and Detroit. @The airline is studying locations of potential hubs in various countries and must consider air mileage between cities and takeoff and landing policies in the various countries. @These cities can be enumerated as the elements of a set, that is: X = (x1, 42,43, x4, x5} = {Omaha, Chicago, Rome, London, Detroit}Crisp Tolerance Relation @Further, suppose we have a tolerance relation, R1, that expresses relationships among these cities: 11000 11001 R=/}00100 00010 01001Crisp Tolerance Relation @This relation is reflexive and symmetric. The graph for this tolerance relation would involve five vertices (five elements in the relation). a) -osoo o-oo 0 0 i 1 R=] 0 0 0 ICrisp Tolerance Relation @The property of reflexivity (diagonal elements equal unity) simply indicates that a city is totally related to itself. C) 11000 Ww 11001 LI R=| 00100 0 — 00010 Ce) 01001 Gomc®)Crisp Tolerance Relation @The property of symmetry might represent proximity: Omaha and Chicago (x1 and x2) are close (in a binary sense) geographically, and Chicago and Detroit (x2 and x5) are close geographically. oS oo R= coo - cof oor o-oo -co-oCrisp Tolerance Relation @This relation, R1, does not have properties of transitivity, for example (x1,92) €Ry (a2,45) € Ri but (x1, x5) ¢ Ri 11000 110 1 Ri=| 00100 00010 o1001Crisp Tolerance Relation @R1 can become an equivalence relation through one (1 < n, where n = 5) composition. 00 | 01 I 0 0 | RroR=] 0 0 I R= o-oo 0 1 coo aes es o> = coroco eo-oceo 1 0Crisp Tolerance Relation @in this matrix that transitivity holds, that is, (x1, x5) € R1, and R is an equivalence relation. RpoRj= os SS Ss a coroFuzzy Equivalence Relations @A fuzzy relation, k, on a single universe X is also a relation from X to X. @lt is a fuzzy equivalence relation if all three of the following properties for matrix relations define it: Reflexivity fg (i, x3) = I. Symmetry eR(Xj, Xj) = UR(X;, Xi). Transitivity —pag(xj,Xj) = Ay and pep, xg) = Ay — M(H, Xe) =A, where A > min[AyFuzzy Tolerance Relations @lf the fuzzy relation ”: satisfies both reflexivity and symmetry then it is called as fuzzy tolerance relation. @A fuzzy tolerance relation can be reformed into fuzzy equivalence relation at most (n-1) compositions. This is given by: RP! = RoR,o---oR:=RFuzzy Tolerance Relations @Consider fuzzy relation is reflexive and symmetric. 1 06 0 02 03 06 1 04 0 08 R=| 0 04 1 0 0 02 0 0 1 05 03 08 0 05 1 @However it is not transitive, e.g., Hp(#1,22) =0.6, pr (x2,25) = 0.8 HR (x1,25) = 0.3 < min (0.8, 0.6).Fuzzy Tolerance Relations @One composition results in the following relation: 1 0.6 0.6 1 R?=RoR=|04 0.4 03° 05 06 08 0.4 0.4 1 0 0.4 0.3 0.5 0 1 0.5 0.6 0.8 0.4 0.5 1 @where transitivity still does not result for example Hr(a1,¢2) =0.6, per (w2,24) = 0.5, pr (21,04) = 0.3 < min (0.6, 0.5).Fuzzy Tolerance Relations @Finally after one or two more compositions, transitivity results 1 0.6 04 05 0.6 06 1 04 05 08 R=|04 04 1 04 04 0.5 0.5 04 1 0.5 0.6 08 04 05 1 @The transitivity is satisfied hence equivalence relation is also satisfied. R3(a,,2