ѧ
JOURNAL OF SOFTWARE
1999 19 3  Vol.19 No.3 1999



֤Լ
ءܼġDavid A.Bell
ժҪԼڴģ벻ȷ.ĳЩ,ͬDempster-Shafer֤ص,Լʹ÷ֻԼ½ƼϽ,Ӷȡ֪ʶʾеĲȷ,֤ʹ֤ݺʵͬһĿ.۱ϵĲ,ʾڵĹϵ,Ƕߵ,ӶΪǵӦƽ·.,֤,֤ݵĻ,Լ,ǷֻĽ,ڡ֤ǷӦڷֻĽ.ͨһ˵شǷ񶨵.
ؼʡ֤,Լ,,ֻ.
ͼšTP18
Evidence Theory and Rough Set Theory
SU Yun-lin 
Department of Computer Science Ji'nan University Guangzhou 510632
GUAN Ji-wenDavid A. Bell
School of Information and Software Engineering University of Ulster at Jordanstown UK
Abstract The rough set theory is introduced to deal with vagueness and uncertainty. In some aspects, the rough set theory overlaps with the Dempster-Shafer theory of evidence, but the rough set theory uses partitions to specify rough sets, lower and upper approximations, and then to capture uncertainty in knowledge representation. In this paper, directing against the discrepancy in the specification between the two theories, the authors explore their relationship in order for ones to understand them and open the way of applying them. In addition, in evidence theory, the basic aperation to combine evidences is the orthogonal sum, while in the rough set theory, the basic operation is the intersection of partition. Therefore, Does the evidence combination correspond to the partition? is the question which may be naturally raised. An example is presented to show that the answer is no.
Key wordsEvidence theory, rough set theory, orthogonal sum, partition.
Dempster-Shafer֤ۣ1֤ͨݺȡ֪ʶʾеĲȷԣ2,Pawlak½Լۣ3ͨʹ÷ֻཻʵͬһĿ.,ֱۺʹöܵձ黶ӭ.,ǰϸΪ.,оͬǰߵĹϵ,ʵҲ.
Ϊʹĸ߿ɶ,ȼԵڵ1нDempster-Shafer֤,֤ݺԼ溯,Լ.
ڵ2,ǽԼ,һȫӼ½ƺϽƣҪʵ,Dempster-Shafer,½Ƶһ.ϽƵһ溯,ϸ֤,ǰûһ47.
3Ϊ,֤а֤ݼӾϵĲ,ͬԼеĻֻĽĹϵ,֤ǷӦڷֻĽ.Ǹһ˵شǷ񶨵.
1֤
ν֤֤ݵһ麯.ЩԼ溯ȵ.ЩǱ˴һһӦ,ÿһκһһṩͬϢ.ΪĵҪ,ǽһһԶ.
ȶһȫU.ǵѡUнеġЩѡӦǻ.
0,1ݱʾ01֮ģ01ʵ.
һd:U0,1һҶ˹ܶȺ,㹫D:.
ͬһصǰҶ˹ܶȺ.һP:2U0,1һҶ˹ܶȺ,й
P1. P()=1.
P2. P(U)=1.
P3. UӼκλ㼯A1,A2,...,An(n1),
.
һM:2U0,1һ,:
M1. m()=0.
M2..
ͬһص.һbel:2U0,1һ,:
B1. bel()=0.
B2. bel(U)=1.
B3. UӼκλ㼯A1,A2,...,An(n1),
.
ڸйЩĳЩҪʵ.
(1) p-d.1,8 dһҶ˹ܶȺ.
(PD),,һҶ˹ܶȺ.
(DP)d(x)=p({x}),xU.
֮,PһҶ˹ܶȺ,(DP)һҶ˹ܶȺ(PD).
(2) bel-m.1,8 mһ,
(BM),,һ.
(MB),.
֮,belһ,(MB)ĺmһ(BM).
(3) ֤ݺҲɽ.,溯plsɽ:
,.
֤ṩıʾĶΪӦṩ˺ܺõ.ͨ½֤ݺ֮ĹϵǱ˴֮Ժ׵ؽת:
,.
ܡ,.
,.
ܡbel(A)=1-pls(),.
pls(A)=1-bel(),,=U-A.
(4) ֤,bel(A)+bel()1.ζŶڲͬAAĳֶ,ǵ.
Ȼ,,bel(A)pls(A).,ǿӼA|bel(A),pls(A)|.,bel(A)ǰ֤֧Aĳ̶,pls(A)=1-bel(A)Aĳ̶.Ҳbel(A)A½,pls(A)AϽ.ǵĲpls(A)-bel(A)ʾʣĺ,ignorance(A)=pls(A)-bel(A).
(5) ֤۲ṩ3ֱʾ֤ݵ֤ݺ,һṩ˽֤ƵǿĲ.
ͷһʵֶڲͬԴ֤ݵ,Բ֤ͬݵ,Dempster-Shafer,ʵݵĲɾԱڼнѡ.
Dempster-ShaferϹ£m1m22Uϵ
.,ɢ m()=0Ң UӼA٧,
m(A)=(1/N)XY=Am1(X)m2(Y).
2U0,1ݵĺmһ.
mΪm1m2ͣΪm1m2.,Ǿ˵m1m2,m1m2ȫͻ.
K=1/NһΪm1m2͹淶.淶K=1/N֮ͻķΧ.
㽻ɼɣ
m1m2=m2m1,
m1(m2m3)=(m1m2)m3.
m1m2ۺϺԼ溯belm1m2=bel1bel2plsm1m2=pls1pls2.ǼΪbel1bel2Լpls1pls2.
2Լۺ֤ݺ
U={u1,u2,...,u|u|},Uϵһȼ۹ϵ.ͨһȼ۹ϵɰUֳɵȼ࣬Ϊ
U/={W1,W2,...,Wj,...,Ww}.
,2Uϵкһ.

ǳһΪڵȼ۹ϵȵ.
U={u1,u2,...,u|u|},ķֱȫUϵĺȵȼ۹ϵȫȼ۹ϵ.U/={{u1},{u2},...,{u|u|}};U/={{u1,u2,...,u|u|}}.
ǶӦڦ,ĵ2Uϵm,m£
(1) mһܶȺ,d:U0,1.2Uеÿµ{u}ÿuU,

(2) mһ,

Uϵһȼ۹ϵӦķֻU/={W1,W2,...,Wj,...,Ww}.ȫUϵһӼǿ(ɵȼ۹ϵӼV)½V-ϽV+£
.
Ӽ,Ϊ
(1) ȼ۹ϵ֧V-,
(2) ȼ۹ϵȺV+-V-,
(3) ȼ۹ϵȾܾU-V+.
ȫUϵһȼ۹ϵ.ǶɦȵV½ƵΪһ2U0,1ݵĺbel£,
.
ȫUϵһȼ۹ϵ.ǶɦȵVϽƵΪһ2U0,1ݵĺbel£,
 .
UΪȫUϵһȼ۹ϵmǶӦڦȵһ.ǣc2uϵ½Ƶ

ǶӦ
B1. bel()=0 |-|=0;
B2. bel(U)=0 |V-|=|U|;
B3. UӼκλ㼯A1,A2,...,An(n1),
,

.
ͬ,2UϵϽƵ

ǶӦ溯. 
3Լ֤
12ȫUϵȼ۹ϵ.ڦ12֮һȻ㡰롱򽻣Ϊɣ.
3.1. ȼ۹ϵ1ͦ2Ľ1ɦ2£u(1ɦ2)vҽu1vڽȫUϵĵȼ۹ϵǿɽĺ͵ݴȺԪʺ͵λԪģҵҽ|U|=||=1.
3.2. Ц1ͦЦ2Uĵȼ۹ϵ1ͦ2֣ǶֵĽ£
Ц1ɦЦ2=Ц1ɦ2.
֤,֤ݵĻ,Լ,ǻֵĽ.ǺȻҪ:֤ݵǷӦڷֻĽ?Ǹ˵شǷ񶨵.
3.1:U={1,2,3,4,5,6},
еȼ۹ϵ1ͦ2£
U/1: {1,6},{2,3},{4,5},
U/2: {1,3,4,5,6},{2}.
ǣ1жӦm1:
 


m1(=0 .
2жӦm2:
 

m2(=0.
ǱȽ.һ棬1ɦ2»֣

Ӧm12Ϊ

һ棬֤УǿԼm1m2£



ǿ,m1m2m12.
һɽ:ڹֵཻཻڹֻ,ڹ͵ཻ֤.ǰʹõǼ򵥵ķֻ,ʹó˷͹淶.Ƚһͬıе,Ƿ,֤ϲӦཻĻ,ߵȻͬ.
4
ԼͬDempster-Shafer֤еĹϵ.ԼڼʵѧϽ,д.ĸ˶ԼеĽ(Ͻͽ½)Dempster-Shafer֤۵֤ݺϸ֤.ǻ˵ԼеķֻĽӦDempster-Shaferеϲ,ӶʹҲȷĽ.
оõӢѧѧ.,1940,ڣҪоΪ˹Ȼܣֲʽϵͳ㷨.ܼ,1934,ڣʿʦ,ҪоΪ˹ܣ㷨Զ.David A. Bell,1938,ʿڣʿʦҪоΪ˹ܣ֤ۣ.
ͨѶϵ:أ510632,ϴѧѧϵ
ߵλأϴѧѧϵݡ510632
ܼ David A.Bell˹شѧԼУϢѧԺӢ
E-mail: tylsu@jnu.edu.cn
ο
1Shafer G. A Mathematical Theory of Evidence. New Jersey: Princeton University Press, 1976
2Grjymala-Busse J W. Managing Uncertainty. North-Holland: Kluwer Academic Publisher Group, 1991
3Pawland Z, Grjymala-Busse J, Slowinski R et al. Rough sets. Communications of ACM, 1995,38(11):8995
4Skowren A. The relationship between the rough set theory and evidence theory. Bulletin of the Polish Academy of Science. Technical Science, 1989,37(1):12,8890
5Skowton A. The rough set theory and evidence theory. Foundamenter Information , 1990. 245262
6Skowron A, Grgymala-Busse J W. From rough set theory to evidence theory. In: Yager R R, Fedijji M, Kacprjyk J eds. Advances in the Dempster-Shafer Theory of Evidennce. New York: John Wiley and Sons, Inc., 1994. 193236
7Skowron A, Stepanish J. Intelligence systems based on rough set approach. Foundations of Computing and Decision Sciences, 1993,18(3,4):343360
8Guan J W, Bell D A. Evidence theory and its applications (volume 1.2). Studies in Computer Science and Artificial Intelligence 7.8. North-Holland: Elsevier Science Publishers, 19911992
1998-03-30ո壩
