如何解决贝叶斯信念网络推断仅返回0和1作为值
我创建了一个贝叶斯信念网络，每次我运行一个cpquery()函数时，即使我应该获得值[0; 1]，结果还是0或1。我尝试将n中的cpquery()增加到相同。任何帮助或见解将不胜感激。
以下是我的程序
dag <- empty.graph(nodes = c("NS","A","S","DM","P","CM","GC","NL","D"))
arc.set <- matrix(c("NS","D","D"),byrow = TRUE,ncol = 2,dimnames = list(NULL,c("from","to")))
arcs(dag) <- arc.set
nodes(dag)
arcs(dag)
plot(dag)
#dag<-model2network("[NS][A][S][DM][P][CM|P][GC|P][NL|NS:A:S][D|DM:CM:GC:NL]")
graphviz.plot(dag,layout = "dot")
graphviz.plot(dag,layout = "circo")
#### CPT
A.lv <- c("[0;5]","]5;10]","]10;15]")
S.lv <- c("Motionless","Moving")
NS.lv <- c("Cat_A","Cat_B","Cat_C","Cat_D","Cat_E","Cat_F","Empty")
P.lv <- c("EXPERT","NOVICE","NONE","EMPTY")
DM.lv <- c('samsung','LGE','HUAWEI','motorola','ZTE','TCL','Sony','Xiaomi',"Others")
GC.lv <- c("]-20;0]","{0}","]0;20]")
CM.lv <-c("Done","None")
NL.lv <-c("[25;55[","[55;85[","[85;115]")
D.lv <- c("Accept","Refuse")
A.prob <- array(c(0.41,0.41,0.18),dim = 3,dimnames = list(A = A.lv))
S.prob <- array(c(0.51,0.49),dim = 2,dimnames = list(S = S.lv))
NS.prob <-array(c(0.1,0.11,0.08,0.06,0.05,0.1,0.5),dim = 7,dimnames = list(NS = NS.lv))
P.prob <- array(c(0.09,0.33,0.53,0.05),dim = 4,dimnames = list(P = P.lv))
DM.prob <-array(c(0.38,0.16,0.02,0.16),dim = 9,dimnames = list(DM = DM.lv))
CM.prob <-array(c(0.02,0.98,0.0002,0.9998,0.0006,0.9994,0.01,0.99),dim=c(2,4),dimnames = list(CM=CM.lv,P=P.lv))
GC.prob <-array(c(0.15,0.73,0.12,0.27,0.63,0.85,0.2,0.78,0.02),dim=c(3,dimnames = list(GC=GC.lv,P=P.lv))
NL.prob <- array(c(0.4,0.57,0.03,0.42,0.55,0.38,0.58,0.04,0.56,0.37,0.59,0.32,0.64,0.36,0.6,0.75,0.18,0.77,0.14,0.8,0.09,0.07,0.82,0.74,0.54,0.43,0.39,0.62,0.69,0.24,0.72,0.23,0.25,0.66,0.7,0.13,0.83,0.71,0.34,0.35,0.61,0.29,0.31,0.21,0.06),dim = c(3,7,2,3),dimnames = list(NL = NL.lv,NS=NS.lv,S = S.lv,A = A.lv))
D.prob <-array(c(0.84,0.99,0.15,##D=A;DM;CM=Y;GC=NV1;NL=NV1
0.77,0.86,0.95,0.87,0.91,0.9,0.44,0.79,##D=A;DM;CM=N;GC=NV1;NL=NV1
0.99,0.67,##D=A;DM;CM=Y;GC=NV2;NL=NV1
0.63,0.4,0.65,0.68,##D=A;DM;CM=N;GC=NV2;NL=NV1
0.99,##D=A;DM;CM=Y;GC=NV3;NL=NV1
0.61,0.28,0.48,0.52,##D=A;DM;CM=N;GC=NV3;NL=NV1
0.58,0.5,##D=A;DM;CM=Y;GC=NV1;NL=NV2
0.76,0.97,0.17,0.76,##D=A;DM;CM=N;GC=NV1;NL=NV2
0.99,##D=A;DM;CM=Y;GC=NV2;NL=NV2
0.69,0.3,##D=A;DM;CM=N;GC=NV2;NL=NV2
0.78,0.22,##D=A;DM;CM=Y;GC=NV3;NL=NV2
0.73,##D=A;DM;CM=N;GC=NV3;NL=NV2
0.99,##D=A;DM;CM=Y;GC=NV1;NL=NV3
0.8,0.89,##D=A;DM;CM=N;GC=NV1;NL=NV3
0.01,##D=A;DM;CM=Y;GC=NV2;NL=NV3
0.75,0.51,0.49,0.26,##D=A;DM;CM=N;GC=NV2;NL=NV3
0.33,##D=A;DM;CM=Y;GC=NV3;NL=NV3
0.71,0.84,0.34 ##D=A;DM;CM=N;GC=NV3;NL=NV3
),dim = c(2,9,3,dimnames = list(D=D.lv,DM=DM.lv,CM=CM.lv,GC=GC.lv,NL=NL.lv))
cpt <- list(A = A.prob,S = S.prob,NS = NS.prob,P = P.prob,DM = DM.prob,CM = CM.prob,GC=GC.prob,NL=NL.prob,D=D.prob)
cpt
bn.global<-custom.fit(dag,cpt)
有人可以帮助我，告诉我哪里出问题了吗？
版权声明：本文内容由互联网用户自发贡献，该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容， 请发送邮件至 dio@foxmail.com 举报，一经查实，本站将立刻删除。