-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdlt.deepX.R
601 lines (546 loc) · 26 KB
/
dlt.deepX.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
row.ab<-dim(dlt)[1]
clu.ab<-dim(dlt)[2]
ab.drf <-c(0,0,0,0,0,0,0)
ab.drfII <-c(0,0,0,0,0,0,0)
ab.bnf <-c(0,0,0,0,0,0,0)
ab.bnfII <-c(0,0,0,0,0,0,0)
ab.xgb <-c(0,0,0,0,0,0,0)
ab.xgbII <-c(0,0,0,0,0,0,0)
ab.bnfAB <-c(0,0,0,0,0,0,0)
ab.glmf <-c(0,0,0,0,0,0,0)
ab.glmfII<-c(0,0,0,0,0,0,0)
ab.svmf <-c(0,0,0,0,0,0,0)
ab.svmfII<-c(0,0,0,0,0,0,0)
ab.knnf <-c(0,0,0,0,0,0,0)
ab.knnfII<-c(0,0,0,0,0,0,0)
s.n<-floor(row.ab/2)
for(i in s.n:row.ab) {
data<-head(dlt,i)
print("#######################################################################################")
print(c(i,row.ab))
count<-s.n
#Randomforest
t.drf <-dlt.DRF(data,count)
ab.drf <-c(ab.drf,t.drf)
t.drfII <-dlt.DRFII(data,count)
ab.drfII<-c(ab.drfII,t.drfII)
#Bayess
t.bnf <-dlt.BNF(data,count)
ab.bnf <-c(ab.bnf,t.bnf)
t.bnfII <-dlt.BNFII(data,count)
ab.bnfII<-c(ab.bnfII,t.bnfII)
#xgBoost
t.xgb <-dlt.XGB(data,count)
ab.xgb <-c(ab.xgb,t.xgb)
t.xgbII <-dlt.XGBII(data,count)
ab.xgbII<-c(ab.xgbII,t.xgbII)
#GLM
t.glmf <-dlt.GLMF(data,count)
ab.glmf <-c(ab.glmf,t.glmf)
t.glmfII <-dlt.GLMFII(data,count)
ab.glmfII<-c(ab.glmfII,t.glmfII)
#SVM
t.svmf <-dlt.SVMF(data,count)
ab.svmf <-c(ab.svmf,t.svmf)
t.svmfII <-dlt.SVMFII(data,count)
ab.svmfII<-c(ab.svmfII,t.svmfII)
#KNN
t.knnf <-dlt.KNNF(data,count)
ab.knnf <-c(ab.knnf,t.knnf)
t.knnfII <-dlt.KNNFII(data,count)
ab.knnfII<-c(ab.knnfII,t.knnfII)
#Bayess AB
t.bnfAB<-dlt.BNab(data,count)
ab.bnfAB<-c(ab.bnfAB,t.bnfAB)
}
#Build matrix
ab.drf.m <-matrix(ab.drf ,ncol = 7,byrow = TRUE)[-1,]
ab.drfII.m<-matrix(ab.drfII,ncol = 7,byrow = TRUE)[-1,]
ab.bnf.m <-matrix(ab.bnf ,ncol = 7,byrow = TRUE)[-1,]
ab.bnfII.m<-matrix(ab.bnfII,ncol = 7,byrow = TRUE)[-1,]
ab.xgb.m <-matrix(ab.xgb ,ncol = 7,byrow = TRUE)[-1,]
ab.xgbII.m<-matrix(ab.xgbII,ncol = 7,byrow = TRUE)[-1,]
ab.glmf.m <-matrix(ab.glmf ,ncol = 7,byrow = TRUE)[-1,]
ab.glmfII.m<-matrix(ab.glmfII,ncol = 7,byrow = TRUE)[-1,]
ab.svmf.m <-matrix(ab.svmf ,ncol = 7,byrow = TRUE)[-1,]
ab.svmfII.m<-matrix(ab.svmfII,ncol = 7,byrow = TRUE)[-1,]
ab.knnf.m <-matrix(ab.knnf ,ncol = 7,byrow = TRUE)[-1,]
ab.knnfII.m<-matrix(ab.knnfII,ncol = 7,byrow = TRUE)[-1,]
ab.bnfAB.m<-matrix(ab.bnfAB,ncol = 7,byrow = TRUE)[-1,]
#Build data.frame
a1.drf <-ab.drf.m [1:(dim(ab.drf.m)[1]-1) ,1]
a2.drf <-ab.drf.m [1:(dim(ab.drf.m)[1]-1) ,2]
a3.drf <-ab.drf.m [1:(dim(ab.drf.m)[1]-1) ,3]
a4.drf <-ab.drf.m [1:(dim(ab.drf.m)[1]-1) ,4]
a5.drf <-ab.drf.m [1:(dim(ab.drf.m)[1]-1) ,5]
b1.drf <-ab.drf.m [1:(dim(ab.drf.m)[1]-1) ,6]
b2.drf <-ab.drf.m [1:(dim(ab.drf.m)[1]-1) ,7]
a1.drfII<-ab.drfII.m[1:(dim(ab.drfII.m)[1]-1),1]
a2.drfII<-ab.drfII.m[1:(dim(ab.drfII.m)[1]-1),2]
a3.drfII<-ab.drfII.m[1:(dim(ab.drfII.m)[1]-1),3]
a4.drfII<-ab.drfII.m[1:(dim(ab.drfII.m)[1]-1),4]
a5.drfII<-ab.drfII.m[1:(dim(ab.drfII.m)[1]-1),5]
b1.drfII<-ab.drfII.m[1:(dim(ab.drfII.m)[1]-1),6]
b2.drfII<-ab.drfII.m[1:(dim(ab.drfII.m)[1]-1),7]
a1.bnf <-ab.bnf.m [1:(dim(ab.bnf.m)[1]-1) ,1]
a2.bnf <-ab.bnf.m [1:(dim(ab.bnf.m)[1]-1) ,2]
a3.bnf <-ab.bnf.m [1:(dim(ab.bnf.m)[1]-1) ,3]
a4.bnf <-ab.bnf.m [1:(dim(ab.bnf.m)[1]-1) ,4]
a5.bnf <-ab.bnf.m [1:(dim(ab.bnf.m)[1]-1) ,5]
b1.bnf <-ab.bnf.m [1:(dim(ab.bnf.m)[1]-1) ,6]
b2.bnf <-ab.bnf.m [1:(dim(ab.bnf.m)[1]-1) ,7]
a1.bnfII<-ab.bnfII.m[1:(dim(ab.bnfII.m)[1]-1),1]
a2.bnfII<-ab.bnfII.m[1:(dim(ab.bnfII.m)[1]-1),2]
a3.bnfII<-ab.bnfII.m[1:(dim(ab.bnfII.m)[1]-1),3]
a4.bnfII<-ab.bnfII.m[1:(dim(ab.bnfII.m)[1]-1),4]
a5.bnfII<-ab.bnfII.m[1:(dim(ab.bnfII.m)[1]-1),5]
b1.bnfII<-ab.bnfII.m[1:(dim(ab.bnfII.m)[1]-1),6]
b2.bnfII<-ab.bnfII.m[1:(dim(ab.bnfII.m)[1]-1),7]
a1.xgb <-ab.xgb.m [1:(dim(ab.xgb.m)[1]-1) ,1]
a2.xgb <-ab.xgb.m [1:(dim(ab.xgb.m)[1]-1) ,2]
a3.xgb <-ab.xgb.m [1:(dim(ab.xgb.m)[1]-1) ,3]
a4.xgb <-ab.xgb.m [1:(dim(ab.xgb.m)[1]-1) ,4]
a5.xgb <-ab.xgb.m [1:(dim(ab.xgb.m)[1]-1) ,5]
b1.xgb <-ab.xgb.m [1:(dim(ab.xgb.m)[1]-1) ,6]
b2.xgb <-ab.xgb.m [1:(dim(ab.xgb.m)[1]-1) ,7]
a1.xgbII<-ab.xgbII.m[1:(dim(ab.xgbII.m)[1]-1),1]
a2.xgbII<-ab.xgbII.m[1:(dim(ab.xgbII.m)[1]-1),2]
a3.xgbII<-ab.xgbII.m[1:(dim(ab.xgbII.m)[1]-1),3]
a4.xgbII<-ab.xgbII.m[1:(dim(ab.xgbII.m)[1]-1),4]
a5.xgbII<-ab.xgbII.m[1:(dim(ab.xgbII.m)[1]-1),5]
b1.xgbII<-ab.xgbII.m[1:(dim(ab.xgbII.m)[1]-1),6]
b2.xgbII<-ab.xgbII.m[1:(dim(ab.xgbII.m)[1]-1),7]
a1.bnfAB<-ab.bnfAB.m[1:(dim(ab.bnfAB.m)[1]-1),1]
a2.bnfAB<-ab.bnfAB.m[1:(dim(ab.bnfAB.m)[1]-1),2]
a3.bnfAB<-ab.bnfAB.m[1:(dim(ab.bnfAB.m)[1]-1),3]
a4.bnfAB<-ab.bnfAB.m[1:(dim(ab.bnfAB.m)[1]-1),4]
a5.bnfAB<-ab.bnfAB.m[1:(dim(ab.bnfAB.m)[1]-1),5]
b1.bnfAB<-ab.bnfAB.m[1:(dim(ab.bnfAB.m)[1]-1),6]
b2.bnfAB<-ab.bnfAB.m[1:(dim(ab.bnfAB.m)[1]-1),7]
a1.glmf <-ab.glmf.m[1:(dim(ab.glmf.m)[1]-1) ,1]
a2.glmf <-ab.glmf.m[1:(dim(ab.glmf.m)[1]-1) ,2]
a3.glmf <-ab.glmf.m[1:(dim(ab.glmf.m)[1]-1) ,3]
a4.glmf <-ab.glmf.m[1:(dim(ab.glmf.m)[1]-1) ,4]
a5.glmf <-ab.glmf.m[1:(dim(ab.glmf.m)[1]-1) ,5]
b1.glmf <-ab.glmf.m[1:(dim(ab.glmf.m)[1]-1) ,6]
b2.glmf <-ab.glmf.m[1:(dim(ab.glmf.m)[1]-1) ,7]
a1.glmfII<-ab.glmfII.m[1:(dim(ab.glmfII.m)[1]-1),1]
a2.glmfII<-ab.glmfII.m[1:(dim(ab.glmfII.m)[1]-1),2]
a3.glmfII<-ab.glmfII.m[1:(dim(ab.glmfII.m)[1]-1),3]
a4.glmfII<-ab.glmfII.m[1:(dim(ab.glmfII.m)[1]-1),4]
a5.glmfII<-ab.glmfII.m[1:(dim(ab.glmfII.m)[1]-1),5]
b1.glmfII<-ab.glmfII.m[1:(dim(ab.glmfII.m)[1]-1),6]
b2.glmfII<-ab.glmfII.m[1:(dim(ab.glmfII.m)[1]-1),7]
a1.svmf <-ab.svmf.m[1:(dim(ab.svmf.m)[1]-1) ,1]
a2.svmf <-ab.svmf.m[1:(dim(ab.svmf.m)[1]-1) ,2]
a3.svmf <-ab.svmf.m[1:(dim(ab.svmf.m)[1]-1) ,3]
a4.svmf <-ab.svmf.m[1:(dim(ab.svmf.m)[1]-1) ,4]
a5.svmf <-ab.svmf.m[1:(dim(ab.svmf.m)[1]-1) ,5]
b1.svmf <-ab.svmf.m[1:(dim(ab.svmf.m)[1]-1) ,6]
b2.svmf <-ab.svmf.m[1:(dim(ab.svmf.m)[1]-1) ,7]
a1.svmfII<-ab.svmfII.m[1:(dim(ab.svmfII.m)[1]-1),1]
a2.svmfII<-ab.svmfII.m[1:(dim(ab.svmfII.m)[1]-1),2]
a3.svmfII<-ab.svmfII.m[1:(dim(ab.svmfII.m)[1]-1),3]
a4.svmfII<-ab.svmfII.m[1:(dim(ab.svmfII.m)[1]-1),4]
a5.svmfII<-ab.svmfII.m[1:(dim(ab.svmfII.m)[1]-1),5]
b1.svmfII<-ab.svmfII.m[1:(dim(ab.svmfII.m)[1]-1),6]
b2.svmfII<-ab.svmfII.m[1:(dim(ab.svmfII.m)[1]-1),7]
a1.knnf <-ab.knnf.m[1:(dim(ab.knnf.m)[1]-1) ,1]
a2.knnf <-ab.knnf.m[1:(dim(ab.knnf.m)[1]-1) ,2]
a3.knnf <-ab.knnf.m[1:(dim(ab.knnf.m)[1]-1) ,3]
a4.knnf <-ab.knnf.m[1:(dim(ab.knnf.m)[1]-1) ,4]
a5.knnf <-ab.knnf.m[1:(dim(ab.knnf.m)[1]-1) ,5]
b1.knnf <-ab.knnf.m[1:(dim(ab.knnf.m)[1]-1) ,6]
b2.knnf <-ab.knnf.m[1:(dim(ab.knnf.m)[1]-1) ,7]
a1.knnfII<-ab.knnfII.m[1:(dim(ab.knnfII.m)[1]-1),1]
a2.knnfII<-ab.knnfII.m[1:(dim(ab.knnfII.m)[1]-1),2]
a3.knnfII<-ab.knnfII.m[1:(dim(ab.knnfII.m)[1]-1),3]
a4.knnfII<-ab.knnfII.m[1:(dim(ab.knnfII.m)[1]-1),4]
a5.knnfII<-ab.knnfII.m[1:(dim(ab.knnfII.m)[1]-1),5]
b1.knnfII<-ab.knnfII.m[1:(dim(ab.knnfII.m)[1]-1),6]
b2.knnfII<-ab.knnfII.m[1:(dim(ab.knnfII.m)[1]-1),7]
result<-dlt[(s.n+1):row.ab,]
a1.r<-result[, 4]
a2.r<-result[, 5]
a3.r<-result[, 6]
a4.r<-result[, 7]
a5.r<-result[, 8]
b1.r<-result[, 9]
b2.r<-result[,10]
#Build training data
train.a1<-data.frame(a1.drf,a1.drfII,a1.xgb,a1.xgbII,
a1.bnf,a1.bnfII,a1.svmf,a1.svmfII,a1.knnf,a1.knnfII,
# a1.glmf,a1.glmfII,
a1.bnfAB,a1.r)
train.a2<-data.frame(
a2.drf,a2.drfII,a2.xgb,a2.xgbII,
a2.bnf,a2.bnfII,a2.svmf,a2.svmfII,a2.knnf,a2.knnfII,
# a2.glmf,a2.glmfII,
a2.bnfAB,a2.r)
train.a3<-data.frame(
a3.drf,a3.drfII,a3.xgb,a3.xgbII,
a3.bnf,a3.bnfII,a3.svmf,a3.svmfII,a3.knnf,a3.knnfII,
# a3.glmf,a3.glmfII,
a3.bnfAB,a3.r)
train.a4<-data.frame(a4.drf,a4.drfII,a4.xgb,a4.xgbII,
a4.bnf,a4.bnfII,a4.svmf,a4.svmfII,a4.knnf,a4.knnfII,
# a4.glmf,a4.glmfII,
a4.bnfAB,a4.r)
train.a5<-data.frame(a5.drf,a5.drfII,a5.xgb,a5.xgbII,
a5.bnf,a5.bnfII,a5.svmf,a5.svmfII,a5.knnf,a5.knnfII,
# a5.glmf,a5.glmfII,
a5.bnfAB,a5.r)
train.b1<-data.frame(b1.drf,b1.drfII,b1.xgb,b1.xgbII,
b1.bnf,b1.bnfII,b1.svmf,b1.svmfII,b1.knnf,b1.knnfII,
# b1.glmf,b1.glmfII,
b1.bnfAB,b1.r)
train.b2<-data.frame(b2.drf,b2.drfII,b2.xgb,b2.xgbII,
b2.bnf,b2.bnfII,b2.svmf,b2.svmfII,b2.knnf,b2.knnfII,
# b2.glmf,b2.glmfII,
b2.bnfAB,b2.r)
#randomForest training
rf.a1 = randomForest(a1.r ~ a1.drf+a1.drfII+a1.xgb+a1.xgbII
+a1.bnf+a1.bnfII+a1.svmf+a1.svmfII+a1.knnf+a1.knnfII
# +a1.glmf+a1.glmfII
+a1.bnfAB,data = train.a1,importance = T,ntrees=1000)
rf.a2 = randomForest(a2.r ~ a2.drf+a2.drfII+a2.xgb+a2.xgbII
+a2.bnf+a2.bnfII+a1.svmf+a2.svmfII+a2.knnf+a2.knnfII
# +a2.glmf+a2.glmfII
+a2.bnfAB,data = train.a2,importance = T,ntrees=1000)
rf.a3 = randomForest(a3.r ~ a3.drf+a3.drfII+a3.xgb+a3.xgbII
+a3.bnf+a3.bnfII+a1.svmf+a3.svmfII+a3.knnf+a3.knnfII
# +a3.glmf+a3.glmfII
+a3.bnfAB,data = train.a3,importance = T,ntrees=1000)
rf.a4 = randomForest(a4.r ~ a4.drf+a4.drfII+a4.xgb+a4.xgbII
+a4.bnf+a4.bnfII+a1.svmf+a4.svmfII+a4.knnf+a4.knnfII
# +a4.glmf+a4.glmfII
+a4.bnfAB,data = train.a4,importance = T,ntrees=1000)
rf.a5 = randomForest(a5.r ~ a5.drf+a5.drfII+a5.xgb+a5.xgbII
+a5.bnf+a5.bnfII+a1.svmf+a5.svmfII+a5.knnf+a5.knnfII
# +a5.glmf+a5.glmfII
+a5.bnfAB,data = train.a5,importance = T,ntrees=1000)
rf.b1 = randomForest(b1.r ~ b1.drf+b1.drfII+b1.xgb+b1.xgbII
+b1.bnf+b1.bnfII+b1.svmf+b1.svmfII+b1.knnf+b1.knnfII
# +b1.glmf+b1.glmfII
+b1.bnfAB,data = train.b1,importance = T,ntrees=1000)
rf.b2 = randomForest(b2.r ~ b2.drf+b2.drfII+b2.xgb+b2.xgbII
+b2.bnf+b2.bnfII+b1.svmf+b2.svmfII+b2.knnf+b2.knnfII
# +b2.glmf+b2.glmfII
+b2.bnfAB,data = train.b2,importance = T,ntrees=1000)
#Bayes training
#A1:
bn.a1<-bnlearn::hc(train.a1)
graphviz.plot(bn.a1, layout = "fdp")
#A2:
bn.a2<-bnlearn::hc(train.a2)
graphviz.plot(bn.a2, layout = "fdp")
#xgboost training
#xgb.a1<-Matrix(as.matrix(train.a1[,1:13]),sparse=T)
#xgb.a2<-Matrix(as.matrix(train.a2[,1:13]),sparse=T)
#xgb.a3<-Matrix(as.matrix(train.a3[,1:13]),sparse=T)
#xgb.a4<-Matrix(as.matrix(train.a4[,1:13]),sparse=T)
#xgb.a5<-Matrix(as.matrix(train.a5[,1:13]),sparse=T)
#xgb.b1<-Matrix(as.matrix(train.b1[,1:13]),sparse=T)
#xgb.b2<-Matrix(as.matrix(train.b2[,1:13]),sparse=T)
xgb.a1<-Matrix(as.matrix(train.a1[,1:11]),sparse=T)
xgb.a2<-Matrix(as.matrix(train.a2[,1:11]),sparse=T)
xgb.a3<-Matrix(as.matrix(train.a3[,1:11]),sparse=T)
xgb.a4<-Matrix(as.matrix(train.a4[,1:11]),sparse=T)
xgb.a5<-Matrix(as.matrix(train.a5[,1:11]),sparse=T)
xgb.b1<-Matrix(as.matrix(train.b1[,1:11]),sparse=T)
xgb.b2<-Matrix(as.matrix(train.b2[,1:11]),sparse=T)
n=300
bst.a1 <- xgboost(data = xgb.a1,label = train.a1$a1.r,nrounds = n,print_every_n = 300L)
bst.a2 <- xgboost(data = xgb.a2,label = train.a2$a2.r,nrounds = n,print_every_n = 300L)
bst.a3 <- xgboost(data = xgb.a3,label = train.a3$a3.r,nrounds = n,print_every_n = 300L)
bst.a4 <- xgboost(data = xgb.a4,label = train.a4$a4.r,nrounds = n,print_every_n = 300L)
bst.a5 <- xgboost(data = xgb.a5,label = train.a5$a5.r,nrounds = n,print_every_n = 300L)
bst.b1 <- xgboost(data = xgb.b1,label = train.b1$b1.r,nrounds = n,print_every_n = 300L)
bst.b2 <- xgboost(data = xgb.b2,label = train.b2$b2.r,nrounds = n,print_every_n = 300L)
#GLM training
#fit_glm.a1 = glmnet(as.matrix(train.a1[,1:13]), train.a1$a1.r, family="gaussian", nlambda=50, alpha=1)
#fit_glm.a2 = glmnet(as.matrix(train.a2[,1:13]), train.a2$a2.r, family="gaussian", nlambda=50, alpha=1)
#fit_glm.a3 = glmnet(as.matrix(train.a3[,1:13]), train.a3$a3.r, family="gaussian", nlambda=50, alpha=1)
#fit_glm.a4 = glmnet(as.matrix(train.a4[,1:13]), train.a4$a4.r, family="gaussian", nlambda=50, alpha=1)
#fit_glm.a5 = glmnet(as.matrix(train.a5[,1:13]), train.a5$a5.r, family="gaussian", nlambda=50, alpha=1)
#fit_glm.b1 = glmnet(as.matrix(train.b1[,1:13]), train.b1$b1.r, family="gaussian", nlambda=50, alpha=1)
#fit_glm.b2 = glmnet(as.matrix(train.b2[,1:13]), train.b1$b1.r, family="gaussian", nlambda=50, alpha=1)
#cvfit_glm.a1 = cv.glmnet(as.matrix(train.a1[,1:13]), train.a1$a1.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
#cvfit_glm.a2 = cv.glmnet(as.matrix(train.a2[,1:13]), train.a2$a2.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
#cvfit_glm.a3 = cv.glmnet(as.matrix(train.a3[,1:13]), train.a3$a3.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
#cvfit_glm.a4 = cv.glmnet(as.matrix(train.a4[,1:13]), train.a4$a4.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
#cvfit_glm.a5 = cv.glmnet(as.matrix(train.a5[,1:13]), train.a5$a5.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
#cvfit_glm.b1 = cv.glmnet(as.matrix(train.b1[,1:13]), train.b1$b1.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
#cvfit_glm.b2 = cv.glmnet(as.matrix(train.b2[,1:13]), train.b2$b2.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
fit_glm.a1 = glmnet(as.matrix(train.a1[,1:11]), train.a1$a1.r, family="gaussian", nlambda=50, alpha=1)
fit_glm.a2 = glmnet(as.matrix(train.a2[,1:11]), train.a2$a2.r, family="gaussian", nlambda=50, alpha=1)
fit_glm.a3 = glmnet(as.matrix(train.a3[,1:11]), train.a3$a3.r, family="gaussian", nlambda=50, alpha=1)
fit_glm.a4 = glmnet(as.matrix(train.a4[,1:11]), train.a4$a4.r, family="gaussian", nlambda=50, alpha=1)
fit_glm.a5 = glmnet(as.matrix(train.a5[,1:11]), train.a5$a5.r, family="gaussian", nlambda=50, alpha=1)
fit_glm.b1 = glmnet(as.matrix(train.b1[,1:11]), train.b1$b1.r, family="gaussian", nlambda=50, alpha=1)
fit_glm.b2 = glmnet(as.matrix(train.b2[,1:11]), train.b1$b1.r, family="gaussian", nlambda=50, alpha=1)
cvfit_glm.a1 = cv.glmnet(as.matrix(train.a1[,1:11]), train.a1$a1.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
cvfit_glm.a2 = cv.glmnet(as.matrix(train.a2[,1:11]), train.a2$a2.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
cvfit_glm.a3 = cv.glmnet(as.matrix(train.a3[,1:11]), train.a3$a3.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
cvfit_glm.a4 = cv.glmnet(as.matrix(train.a4[,1:11]), train.a4$a4.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
cvfit_glm.a5 = cv.glmnet(as.matrix(train.a5[,1:11]), train.a5$a5.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
cvfit_glm.b1 = cv.glmnet(as.matrix(train.b1[,1:11]), train.b1$b1.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
cvfit_glm.b2 = cv.glmnet(as.matrix(train.b2[,1:11]), train.b2$b2.r, family = "gaussian", type.measure = "mse",parallel = TRUE)
#KNN
#Build KNN model
#fit_knn.a1 <- knnreg(train.a1[,1:13], train.a1$a1.r, k = 10)
#fit_knn.a2 <- knnreg(train.a2[,1:13], train.a2$a2.r, k = 10)
#fit_knn.a3 <- knnreg(train.a3[,1:13], train.a3$a3.r, k = 10)
#fit_knn.a4 <- knnreg(train.a4[,1:13], train.a4$a4.r, k = 10)
#fit_knn.a5 <- knnreg(train.a5[,1:13], train.a5$a5.r, k = 10)
#fit_knn.b1 <- knnreg(train.b1[,1:13], train.b1$b1.r, k = 10)
#fit_knn.b2 <- knnreg(train.b2[,1:13], train.b2$b2.r, k = 10)
fit_knn.a1 <- knnreg(train.a1[,1:11], train.a1$a1.r, k = 10)
fit_knn.a2 <- knnreg(train.a2[,1:11], train.a2$a2.r, k = 10)
fit_knn.a3 <- knnreg(train.a3[,1:11], train.a3$a3.r, k = 10)
fit_knn.a4 <- knnreg(train.a4[,1:11], train.a4$a4.r, k = 10)
fit_knn.a5 <- knnreg(train.a5[,1:11], train.a5$a5.r, k = 10)
fit_knn.b1 <- knnreg(train.b1[,1:11], train.b1$b1.r, k = 10)
fit_knn.b2 <- knnreg(train.b2[,1:11], train.b2$b2.r, k = 10)
#SVM training
fit_svm.a1=svm(a1.r ~ a1.drf+a1.drfII+a1.xgb+a1.xgbII
+a1.bnf+a1.bnfII+a1.svmf+a1.svmfII+a1.knnf+a1.knnfII
# +a1.glmf+a1.glmfII
+a1.bnfAB,data = train.a1)
fit_svm.a2 = svm(a2.r ~ a2.drf+a2.drfII+a2.xgb+a2.xgbII
+a2.bnf+a2.bnfII+a1.svmf+a2.svmfII+a2.knnf+a2.knnfII
# +a2.glmf+a2.glmfII
+a2.bnfAB,data = train.a2)
fit_svm.a3 = svm(a3.r ~ a3.drf+a3.drfII+a3.xgb+a3.xgbII
+a3.bnf+a3.bnfII+a1.svmf+a3.svmfII+a3.knnf+a3.knnfII
# +a3.glmf+a3.glmfII
+a3.bnfAB,data = train.a3)
fit_svm.a4 = svm(a4.r ~ a4.drf+a4.drfII+a4.xgb+a4.xgbII
+a4.bnf+a4.bnfII+a1.svmf+a4.svmfII+a4.knnf+a4.knnfII
# +a4.glmf+a4.glmfII
+a4.bnfAB,data = train.a4)
fit_svm.a5 = svm(a5.r ~ a5.drf+a5.drfII+a5.xgb+a5.xgbII
+a5.bnf+a5.bnfII+a1.svmf+a5.svmfII+a5.knnf+a5.knnfII
# +a5.glmf+a5.glmfII
+a5.bnfAB,data = train.a5)
fit_svm.b1 = svm(b1.r ~ b1.drf+b1.drfII+b1.xgb+b1.xgbII
+b1.bnf+b1.bnfII+b1.svmf+b1.svmfII+b1.knnf+b1.knnfII
# +b1.glmf+b1.glmfII
+b1.bnfAB,data = train.b1)
fit_svm.b2 = svm(b2.r ~ b2.drf+b2.drfII+b2.xgb+b2.xgbII
+b2.bnf+b2.bnfII+b1.svmf+b2.svmfII+b2.knnf+b2.knnfII
# +b2.glmf+b2.glmfII
+b2.bnfAB,data = train.b2)
#Build predict data
a1.drf <-tail(ab.drf.m,1)[ 1]
a2.drf <-tail(ab.drf.m,1)[ 2]
a3.drf <-tail(ab.drf.m,1)[ 3]
a4.drf <-tail(ab.drf.m,1)[ 4]
a5.drf <-tail(ab.drf.m,1)[ 5]
b1.drf <-tail(ab.drf.m,1)[ 6]
b2.drf <-tail(ab.drf.m,1)[ 7]
a1.drfII<-tail(ab.drfII.m,1)[1]
a2.drfII<-tail(ab.drfII.m,1)[2]
a3.drfII<-tail(ab.drfII.m,1)[3]
a4.drfII<-tail(ab.drfII.m,1)[4]
a5.drfII<-tail(ab.drfII.m,1)[5]
b1.drfII<-tail(ab.drfII.m,1)[6]
b2.drfII<-tail(ab.drfII.m,1)[7]
a1.bnf <-tail(ab.bnf.m,1)[ 1]
a2.bnf <-tail(ab.bnf.m,1)[ 2]
a3.bnf <-tail(ab.bnf.m,1)[ 3]
a4.bnf <-tail(ab.bnf.m,1)[ 4]
a5.bnf <-tail(ab.bnf.m,1)[ 5]
b1.bnf <-tail(ab.bnf.m,1)[ 6]
b2.bnf <-tail(ab.bnf.m,1)[ 7]
a1.bnfII<-tail(ab.bnfII.m,1)[1]
a2.bnfII<-tail(ab.bnfII.m,1)[2]
a3.bnfII<-tail(ab.bnfII.m,1)[3]
a4.bnfII<-tail(ab.bnfII.m,1)[4]
a5.bnfII<-tail(ab.bnfII.m,1)[5]
b1.bnfII<-tail(ab.bnfII.m,1)[6]
b2.bnfII<-tail(ab.bnfII.m,1)[7]
a1.xgb <-tail(ab.xgb.m,1)[ 1]
a2.xgb <-tail(ab.xgb.m,1)[ 2]
a3.xgb <-tail(ab.xgb.m,1)[ 3]
a4.xgb <-tail(ab.xgb.m,1)[ 4]
a5.xgb <-tail(ab.xgb.m,1)[ 5]
b1.xgb <-tail(ab.xgb.m,1)[ 6]
b2.xgb <-tail(ab.xgb.m,1)[ 7]
a1.xgbII<-tail(ab.xgbII.m,1)[1]
a2.xgbII<-tail(ab.xgbII.m,1)[2]
a3.xgbII<-tail(ab.xgbII.m,1)[3]
a4.xgbII<-tail(ab.xgbII.m,1)[4]
a5.xgbII<-tail(ab.xgbII.m,1)[5]
b1.xgbII<-tail(ab.xgbII.m,1)[6]
b2.xgbII<-tail(ab.xgbII.m,1)[7]
a1.bnfAB<-tail(ab.bnfAB.m,1)[1]
a2.bnfAB<-tail(ab.bnfAB.m,1)[2]
a3.bnfAB<-tail(ab.bnfAB.m,1)[3]
a4.bnfAB<-tail(ab.bnfAB.m,1)[4]
a5.bnfAB<-tail(ab.bnfAB.m,1)[5]
b1.bnfAB<-tail(ab.bnfAB.m,1)[6]
b2.bnfAB<-tail(ab.bnfAB.m,1)[7]
a1.glmf <-tail(ab.glmf.m,1)[1]
a2.glmf <-tail(ab.glmf.m,1)[2]
a3.glmf <-tail(ab.glmf.m,1)[3]
a4.glmf <-tail(ab.glmf.m,1)[4]
a5.glmf <-tail(ab.glmf.m,1)[5]
b1.glmf <-tail(ab.glmf.m,1)[6]
b2.glmf <-tail(ab.glmf.m,1)[7]
a1.glmfII<-tail(ab.glmfII.m,1)[1]
a2.glmfII<-tail(ab.glmfII.m,1)[2]
a3.glmfII<-tail(ab.glmfII.m,1)[3]
a4.glmfII<-tail(ab.glmfII.m,1)[4]
a5.glmfII<-tail(ab.glmfII.m,1)[5]
b1.glmfII<-tail(ab.glmfII.m,1)[6]
b2.glmfII<-tail(ab.glmfII.m,1)[7]
a1.svmf <-tail(ab.svmf.m,1)[1]
a2.svmf <-tail(ab.svmf.m,1)[2]
a3.svmf <-tail(ab.svmf.m,1)[3]
a4.svmf <-tail(ab.svmf.m,1)[4]
a5.svmf <-tail(ab.svmf.m,1)[5]
b1.svmf <-tail(ab.svmf.m,1)[6]
b2.svmf <-tail(ab.svmf.m,1)[7]
a1.svmfII<-tail(ab.svmfII.m,1)[1]
a2.svmfII<-tail(ab.svmfII.m,1)[2]
a3.svmfII<-tail(ab.svmfII.m,1)[3]
a4.svmfII<-tail(ab.svmfII.m,1)[4]
a5.svmfII<-tail(ab.svmfII.m,1)[5]
b1.svmfII<-tail(ab.svmfII.m,1)[6]
b2.svmfII<-tail(ab.svmfII.m,1)[7]
a1.knnf <-tail(ab.knnf.m,1)[1]
a2.knnf <-tail(ab.knnf.m,1)[2]
a3.knnf <-tail(ab.knnf.m,1)[3]
a4.knnf <-tail(ab.knnf.m,1)[4]
a5.knnf <-tail(ab.knnf.m,1)[5]
b1.knnf <-tail(ab.knnf.m,1)[6]
b2.knnf <-tail(ab.knnf.m,1)[7]
a1.knnfII<-tail(ab.knnfII.m,1)[1]
a2.knnfII<-tail(ab.knnfII.m,1)[2]
a3.knnfII<-tail(ab.knnfII.m,1)[3]
a4.knnfII<-tail(ab.knnfII.m,1)[4]
a5.knnfII<-tail(ab.knnfII.m,1)[5]
b1.knnfII<-tail(ab.knnfII.m,1)[6]
b2.knnfII<-tail(ab.knnfII.m,1)[7]
test.a1<-data.frame(a1.drf,a1.drfII,a1.xgb,a1.xgbII,
a1.bnf,a1.bnfII,a1.svmf,a1.svmfII,a1.knnf,a1.knnfII,
# a1.glmf,a1.glmfII,
a1.bnfAB)
test.a2<-data.frame(a2.drf,a2.drfII,a2.xgb,a2.xgbII,
a2.bnf,a2.bnfII,a2.svmf,a2.svmfII,a2.knnf,a2.knnfII,
# a2.glmf,a2.glmfII,
a2.bnfAB)
test.a3<-data.frame(a3.drf,a3.drfII,a3.xgb,a3.xgbII,
a3.bnf,a3.bnfII,a3.svmf,a3.svmfII,a3.knnf,a3.knnfII,
# a3.glmf,a3.glmfII,
a3.bnfAB)
test.a4<-data.frame(a4.drf,a4.drfII,a4.xgb,a4.xgbII,
a4.bnf,a4.bnfII,a4.svmf,a4.svmfII,a4.knnf,a4.knnfII,
# a4.glmf,a4.glmfII,
a4.bnfAB)
test.a5<-data.frame(a5.drf,a5.drfII,a5.xgb,a5.xgbII,
a5.bnf,a5.bnfII,a5.svmf,a5.svmfII,a5.knnf,a5.knnfII,
# a5.glmf,a5.glmfII,
a5.bnfAB)
test.b1<-data.frame(b1.drf,b1.drfII,b1.xgb,b1.xgbII,
b1.bnf,b1.bnfII,b1.svmf,b1.svmfII,b1.knnf,b1.knnfII,
# b1.glmf,b1.glmfII,
b1.bnfAB)
test.b2<-data.frame(b2.drf,b2.drfII,b2.xgb,b2.xgbII,
b2.bnf,b2.bnfII,b2.svmf,b2.svmfII,b2.knnf,b2.knnfII,
# b2.glmf,b2.glmfII,
b2.bnfAB)
#Bayes predict
#p.a1 <-predict(fit_bn.a1,round(test.a1),node = "a1.r")
#randomForest predict
p.a1 = predict(rf.a1,test.a1)
p.a2 = predict(rf.a2,test.a2)
p.a3 = predict(rf.a3,test.a3)
p.a4 = predict(rf.a4,test.a4)
p.a5 = predict(rf.a5,test.a5)
p.b1 = predict(rf.b1,test.b1)
p.b2 = predict(rf.b2,test.b2)
#Result
r.rdf<-c(p.a1,p.a2,p.a3,p.a4,p.a5,p.b1,p.b2)
#xgboostpredict
#t.a1<-Matrix(as.matrix(test.a1[,1:13]),sparse=T)
#t.a2<-Matrix(as.matrix(test.a2[,1:13]),sparse=T)
#t.a3<-Matrix(as.matrix(test.a3[,1:13]),sparse=T)
#t.a4<-Matrix(as.matrix(test.a4[,1:13]),sparse=T)
#t.a5<-Matrix(as.matrix(test.a5[,1:13]),sparse=T)
#t.b1<-Matrix(as.matrix(test.b1[,1:13]),sparse=T)
#t.b2<-Matrix(as.matrix(test.b2[,1:13]),sparse=T)
t.a1<-Matrix(as.matrix(test.a1[,1:11]),sparse=T)
t.a2<-Matrix(as.matrix(test.a2[,1:11]),sparse=T)
t.a3<-Matrix(as.matrix(test.a3[,1:11]),sparse=T)
t.a4<-Matrix(as.matrix(test.a4[,1:11]),sparse=T)
t.a5<-Matrix(as.matrix(test.a5[,1:11]),sparse=T)
t.b1<-Matrix(as.matrix(test.b1[,1:11]),sparse=T)
t.b2<-Matrix(as.matrix(test.b2[,1:11]),sparse=T)
p.a1 <- predict(object = bst.a1,newdata = t(t.a1))
p.a2 <- predict(object = bst.a2,newdata = t(t.a2))
p.a3 <- predict(object = bst.a3,newdata = t(t.a3))
p.a4 <- predict(object = bst.a4,newdata = t(t.a4))
p.a5 <- predict(object = bst.a5,newdata = t(t.a5))
p.b1 <- predict(object = bst.b1,newdata = t(t.b1))
p.b2 <- predict(object = bst.b2,newdata = t(t.b2))
#Result
result.a<-c(mean(p.a1),mean(p.a2),mean(p.a3),mean(p.a4),mean(p.a5))
r.xgb<-c(sort(result.a),mean(p.b1),mean(p.b2))
#GLM test suit
#p.a1<-predict(cvfit_glm.a1, newx=as.matrix(test.a1[1:13]),s="lambda.1se")
#p.a2<-predict(cvfit_glm.a2, newx=as.matrix(test.a2[1:13]),s="lambda.1se")
#p.a3<-predict(cvfit_glm.a3, newx=as.matrix(test.a3[1:13]),s="lambda.1se")
#p.a4<-predict(cvfit_glm.a4, newx=as.matrix(test.a4[1:13]),s="lambda.1se")
#p.a5<-predict(cvfit_glm.a5, newx=as.matrix(test.a5[1:13]),s="lambda.1se")
#p.b1<-predict(cvfit_glm.b1, newx=as.matrix(test.b1[1:13]),s="lambda.1se")
#p.b2<-predict(cvfit_glm.b2, newx=as.matrix(test.b2[1:13]),s="lambda.1se")
p.a1<-predict(cvfit_glm.a1, newx=as.matrix(test.a1[1:11]),s="lambda.1se")
p.a2<-predict(cvfit_glm.a2, newx=as.matrix(test.a2[1:11]),s="lambda.1se")
p.a3<-predict(cvfit_glm.a3, newx=as.matrix(test.a3[1:11]),s="lambda.1se")
p.a4<-predict(cvfit_glm.a4, newx=as.matrix(test.a4[1:11]),s="lambda.1se")
p.a5<-predict(cvfit_glm.a5, newx=as.matrix(test.a5[1:11]),s="lambda.1se")
p.b1<-predict(cvfit_glm.b1, newx=as.matrix(test.b1[1:11]),s="lambda.1se")
p.b2<-predict(cvfit_glm.b2, newx=as.matrix(test.b2[1:11]),s="lambda.1se")
#Result
r.glm<-c(p.a1,p.a2,p.a3,p.a4,p.a5,p.b1,p.b2)
#KNN test suit
p.a1 <- predict(fit_knn.a1, test.a1)
p.a2 <- predict(fit_knn.a2, test.a2)
p.a3 <- predict(fit_knn.a3, test.a3)
p.a4 <- predict(fit_knn.a4, test.a4)
p.a5 <- predict(fit_knn.a5, test.a5)
p.b1 <- predict(fit_knn.b1, test.b1)
p.b2 <- predict(fit_knn.b2, test.b2)
#Result
r.knn<-c(p.a1,p.a2,p.a3,p.a4,p.a5,p.b1,p.b2)
#SVM test suit
p.a1 <- predict(object = fit_svm.a1,newdata = test.a1)
p.a2 <- predict(object = fit_svm.a2,newdata = test.a2)
p.a3 <- predict(object = fit_svm.a3,newdata = test.a3)
p.a4 <- predict(object = fit_svm.a4,newdata = test.a4)
p.a5 <- predict(object = fit_svm.a5,newdata = test.a5)
p.b1 <- predict(object = fit_svm.b1,newdata = test.b1)
p.b2 <- predict(object = fit_svm.b2,newdata = test.b2)
#Reault
r.svm<-c(p.a1,p.a2,p.a3,p.a4,p.a5,p.b1,p.b2)
drf<-t.drf
drfII<-t.drfII
bnf<-t.bnf
bnII<-t.bnfII
xgb<-t.xgb
xgbII<-t.xgbII
glmf<-t.glmf
glmfII<-t.glmfII
svmf<-t.svmf
svmfII<-t.svmfII
knn<-t.knnf
knnII<-t.knnfII
bnAB<-t.bnfAB
AB.final<-data.frame(drf,drfII,bnf,bnII,xgb,xgbII,glmf,glmfII,svmf,svmfII,knn,knnII,bnAB)
floor(r.rdf)
floor(r.svm)
floor(r.glm)
floor(r.xgb)
floor(r.knn)