All convergence scores per package:algorithm sored by minimum RMSE
|
Input parameter
|
RMSE Score
|
Other score
|
Package
|
Algorithm
|
Input format
|
Maxit
|
Learn.rate
|
median
|
D51
|
MAE
|
WAE
|
nlsr
|
- NashLM
|
full fmla & data
|
200
|
N/A
|
3
|
16
|
3
|
6
|
rminer
|
- nnet_optim(BFGS)
|
fmla & data
|
200
|
N/A
|
1
|
6
|
1
|
1
|
nnet
|
- optim (BFGS)
|
x & y
|
200
|
N/A
|
2
|
17
|
2
|
3
|
validann
|
- optim(BFGS)
|
x & y
|
200
|
N/A
|
4
|
10
|
4
|
5
|
- optim(CG)
|
x & y
|
1000
|
N/A
|
6
|
10
|
5
|
4
|
- optim(L-BFGS-B)
|
x & y
|
200
|
N/A
|
13
|
30
|
14
|
13
|
- optim(Nelder-Mead)
|
x & y
|
10000
|
N/A
|
44
|
45
|
46
|
42
|
- optim(SANN)
|
x & y
|
1000
|
N/A
|
53
|
51
|
56
|
55
|
MachineShop
|
- nnet_optim(BFGS)
|
fmla & data
|
200
|
N/A
|
9
|
22
|
9
|
7
|
traineR
|
- nnet_optim(BFGS)
|
fmla & data
|
200
|
N/A
|
5
|
15
|
6
|
2
|
radiant.model
|
- nnet_optim(BFGS)
|
“y” & data
|
200
|
N/A
|
8
|
32
|
12
|
10
|
monmlp
|
- optimx(BFGS)
|
x & y
|
200
|
N/A
|
10
|
18
|
9
|
11
|
- optimx(Nelder-Mead)
|
x & y
|
10000
|
N/A
|
47
|
45
|
44
|
47
|
CaDENCE
|
- optim(BFGS)
|
x & y
|
200
|
N/A
|
28
|
48
|
21
|
40
|
- Rprop
|
x & y
|
1000
|
0.01
|
54
|
60
|
52
|
58
|
- pso_psoptim
|
x & y
|
1000
|
N/A
|
56
|
56
|
54
|
56
|
h2o
|
- first-order
|
“y” & data
|
10000
|
0.01
|
7
|
7
|
8
|
8
|
EnsembleBase
|
- nnet_optim(BFGS)
|
x & y
|
200
|
N/A
|
15
|
34
|
15
|
15
|
caret
|
- avNNet_nnet_optim(BFGS)
|
x & y
|
200
|
N/A
|
10
|
21
|
11
|
9
|
brnn
|
- Gauss-Newton
|
x & y
|
200
|
N/A
|
12
|
9
|
13
|
12
|
qrnn
|
- nlm()
|
x & y
|
200
|
N/A
|
14
|
25
|
7
|
36
|
RSNNS
|
- Rprop
|
x & y
|
1000
|
N/A
|
23
|
52
|
25
|
28
|
- SCG
|
x & y
|
1000
|
N/A
|
17
|
26
|
18
|
19
|
- Std_Backpropagation
|
x & y
|
1000
|
0.1
|
32
|
31
|
31
|
36
|
- BackpropChunk
|
x & y
|
1000
|
N/A
|
34
|
41
|
32
|
34
|
- BackpropMomentum
|
x & y
|
1000
|
N/A
|
35
|
39
|
35
|
30
|
- BackpropWeightDecay
|
x & y
|
1000
|
N/A
|
30
|
43
|
33
|
31
|
- BackpropBatch
|
x & y
|
10000
|
0.1
|
48
|
27
|
50
|
48
|
- Quickprop
|
x & y
|
10000
|
N/A
|
58
|
36
|
58
|
57
|
automl
|
- trainwgrad_adam
|
x & y
|
1000
|
0.01
|
20
|
35
|
16
|
20
|
- trainwgrad_RMSprop
|
x & y
|
1000
|
0.01
|
31
|
50
|
29
|
39
|
- trainwpso
|
x & y
|
1000
|
N/A
|
41
|
49
|
41
|
38
|
deepnet
|
- BP
|
x & y
|
1000
|
0.8
|
18
|
38
|
24
|
17
|
neuralnet
|
- rprop+
|
fmla & data
|
100000
|
N/A
|
23
|
40
|
23
|
24
|
- rprop-
|
fmla & data
|
100000
|
N/A
|
21
|
42
|
21
|
18
|
- slr
|
fmla & data
|
100000
|
N/A
|
39
|
37
|
39
|
46
|
- sag
|
fmla & data
|
100000
|
N/A
|
49
|
59
|
47
|
52
|
- backprop
|
fmla & data
|
100000
|
0.001
|
51
|
10
|
49
|
45
|
keras
|
- adamax
|
x & y
|
10000
|
0.1
|
18
|
20
|
20
|
16
|
- adam
|
x & y
|
10000
|
0.1
|
28
|
44
|
30
|
25
|
- nadam
|
x & y
|
10000
|
0.1
|
39
|
58
|
40
|
41
|
- adagrad
|
x & y
|
10000
|
0.1
|
43
|
53
|
42
|
35
|
- adadelta
|
x & y
|
10000
|
0.1
|
35
|
19
|
34
|
33
|
- sgd
|
x & y
|
10000
|
0.1
|
45
|
47
|
45
|
43
|
- rmsprop
|
x & y
|
10000
|
0.1
|
55
|
57
|
55
|
54
|
AMORE
|
- ADAPTgdwm
|
x & y
|
1000
|
0.01
|
22
|
29
|
16
|
26
|
- ADAPTgd
|
x & y
|
1000
|
0.01
|
25
|
8
|
26
|
21
|
- BATCHgdwm
|
x & y
|
10000
|
0.1
|
33
|
14
|
37
|
27
|
- BATCHgd
|
x & y
|
10000
|
0.1
|
38
|
24
|
42
|
31
|
minpack.lm
|
- Levenberg-Marquardt
|
full fmla & data
|
200
|
N/A
|
16
|
5
|
19
|
14
|
ANN2
|
- rmsprop
|
x & y
|
1000
|
0.01
|
25
|
33
|
27
|
23
|
- adam
|
x & y
|
1000
|
0.01
|
27
|
27
|
28
|
21
|
- sgd
|
x & y
|
1000
|
0.01
|
37
|
22
|
36
|
29
|
deepdive
|
- adam
|
x & y
|
10000
|
0.4
|
42
|
1
|
38
|
44
|
- rmsProp
|
x & y
|
1000
|
0.8
|
46
|
4
|
48
|
50
|
- momentum
|
x & y
|
1000
|
0.8
|
52
|
3
|
53
|
51
|
- gradientDescent
|
x & y
|
10000
|
0.8
|
57
|
2
|
57
|
53
|
snnR
|
- SemiSmoothNewton
|
x & y
|
200
|
N/A
|
49
|
13
|
50
|
48
|
elmNNRcpp
|
- ELM
|
x & y
|
|
N/A
|
59
|
55
|
59
|
59
|
ELMR
|
- ELM
|
fmla & data
|
|
N/A
|
60
|
53
|
60
|
60
|
Packages Tested per Multivariate Dataset
mDette
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.128
|
0.4391
|
0.4564
|
0.0173
|
0.3246
|
2.0005
|
1
|
train_ADAPTgdwm
|
0.184
|
0.3972
|
0.4012
|
0.0040
|
0.3084
|
1.7312
|
2
|
train_BATCHgd
|
1.870
|
1.8688
|
1.8999
|
0.0311
|
1.5158
|
8.6487
|
3
|
train_BATCHgdwm
|
1.862
|
1.8586
|
1.9806
|
0.1220
|
1.4990
|
11.2445
|
4
|
ANN2
|
neuralnetwork_adam
|
0.218
|
1.7980
|
2.0396
|
0.2416
|
1.5178
|
11.5812
|
5
|
neuralnetwork_rmsprop
|
0.206
|
1.9463
|
2.0761
|
0.1298
|
1.5240
|
12.6858
|
6
|
neuralnetwork_sgd
|
0.204
|
1.2208
|
2.0228
|
0.8020
|
1.4953
|
8.6218
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
9.584
|
0.4255
|
0.6160
|
0.1905
|
0.4710
|
3.2585
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
8.632
|
0.4821
|
0.6996
|
0.2175
|
0.5006
|
3.8172
|
9
|
automl_train_manual_trainwpso
|
13.696
|
2.7275
|
4.9634
|
2.2359
|
3.7904
|
24.2831
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.216
|
0.4578
|
1.9537
|
1.4959
|
1.4572
|
11.8945
|
11
|
CaDENCE
|
cadence.fit_optim
|
7.072
|
0.3277
|
2.5664
|
2.2387
|
1.2936
|
17.3208
|
12
|
cadence.fit_psoptim
|
11.258
|
3.1663
|
3.6338
|
0.4675
|
2.1362
|
22.3798
|
13
|
cadence.fit_Rprop
|
17.178
|
4.6664
|
5.7488
|
1.0824
|
3.4794
|
31.0108
|
14
|
caret
|
avNNet_none
|
0.252
|
0.3175
|
0.3514
|
0.0339
|
0.2681
|
1.8536
|
15
|
deepdive
|
deepnet_adam
|
0.738
|
3.0971
|
3.0971
|
0.0000
|
2.0640
|
18.6373
|
16
|
deepnet_gradientDescent
|
7.266
|
4.4310
|
4.4310
|
0.0000
|
3.2628
|
20.7622
|
17
|
deepnet_momentum
|
7.434
|
4.1990
|
4.1990
|
0.0000
|
3.1011
|
18.5512
|
18
|
deepnet_rmsProp
|
0.758
|
2.7205
|
2.7205
|
0.0000
|
1.8705
|
16.1780
|
19
|
deepnet
|
nn.train_BP
|
0.648
|
0.5308
|
0.6403
|
0.1095
|
0.5135
|
2.7237
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.004
|
7.3193
|
7.6899
|
0.3706
|
5.9574
|
32.3344
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.018
|
6.3469
|
7.2310
|
0.8841
|
5.5344
|
32.0052
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.026
|
0.8770
|
13.9426
|
13.0656
|
11.3013
|
47.5398
|
23
|
h2o
|
h2o.deeplearning_first-order
|
6.274
|
0.3696
|
0.3789
|
0.0093
|
0.2948
|
1.3228
|
24
|
keras
|
fit_adadelta
|
29.372
|
2.0733
|
2.3080
|
0.2347
|
1.5890
|
13.7080
|
25
|
fit_adagrad
|
18.384
|
1.5412
|
2.2114
|
0.6702
|
1.5982
|
12.7204
|
26
|
fit_adam
|
2.068
|
0.7615
|
1.0487
|
0.2872
|
0.7949
|
6.3699
|
27
|
fit_adamax
|
4.386
|
0.6492
|
0.6952
|
0.0460
|
0.5462
|
4.1959
|
28
|
fit_nadam
|
3.422
|
1.0271
|
1.2485
|
0.2214
|
0.9787
|
4.9790
|
29
|
fit_rmsprop
|
1.836
|
2.6780
|
3.2516
|
0.5736
|
2.3382
|
16.3052
|
30
|
fit_sgd
|
8.816
|
0.5726
|
2.3026
|
1.7300
|
1.6878
|
10.2998
|
31
|
MachineShop
|
fit_none
|
0.076
|
0.2570
|
1.2314
|
0.9744
|
0.9854
|
8.0327
|
32
|
minpack.lm
|
nlsLM_none
|
0.242
|
0.6081
|
0.6081
|
0.0000
|
0.4989
|
1.9776
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.298
|
0.3732
|
0.4512
|
0.0780
|
0.3380
|
1.8359
|
34
|
monmlp.fit_Nelder-Mead
|
1.100
|
3.0247
|
3.4557
|
0.4310
|
2.5277
|
18.0917
|
35
|
neuralnet
|
neuralnet_backprop
|
14.200
|
8.1656
|
8.1656
|
0.0000
|
6.5262
|
36.2385
|
36
|
neuralnet_rprop-
|
6.318
|
0.5338
|
2.0473
|
1.5135
|
1.4437
|
12.5391
|
37
|
neuralnet_rprop+
|
3.836
|
0.4859
|
0.5467
|
0.0608
|
0.4149
|
2.3410
|
38
|
neuralnet_sag
|
12.916
|
2.1196
|
8.1656
|
6.0460
|
6.5262
|
36.2385
|
39
|
neuralnet_slr
|
6.914
|
0.5494
|
0.5688
|
0.0194
|
0.4293
|
2.4012
|
40
|
nlsr
|
nlxb_none
|
0.522
|
0.1400
|
0.4500
|
0.3100
|
0.3497
|
2.7841
|
41
|
nnet
|
nnet_none
|
0.078
|
0.2650
|
0.4735
|
0.2085
|
0.3557
|
2.0121
|
42
|
qrnn
|
qrnn.fit_none
|
0.518
|
0.3632
|
0.7514
|
0.3882
|
0.4482
|
6.6249
|
43
|
radiant.model
|
nn_none
|
0.112
|
0.2621
|
0.5412
|
0.2791
|
0.4096
|
2.1475
|
44
|
rminer
|
fit_none
|
0.248
|
0.2335
|
0.3147
|
0.0812
|
0.2456
|
1.2905
|
45
|
RSNNS
|
mlp_BackpropBatch
|
6.752
|
1.9746
|
2.0170
|
0.0424
|
1.5451
|
10.0256
|
46
|
mlp_BackpropChunk
|
0.702
|
0.5892
|
0.7126
|
0.1234
|
0.5252
|
2.8993
|
47
|
mlp_BackpropMomentum
|
0.688
|
0.6547
|
0.7744
|
0.1197
|
0.5909
|
3.1612
|
48
|
mlp_BackpropWeightDecay
|
0.654
|
0.6328
|
0.7698
|
0.1370
|
0.5856
|
3.0364
|
49
|
mlp_Quickprop
|
7.460
|
7.1667
|
7.3190
|
0.1523
|
6.0055
|
29.6111
|
50
|
mlp_Rprop
|
0.692
|
0.7757
|
1.2553
|
0.4796
|
0.9246
|
7.6985
|
51
|
mlp_SCG
|
1.156
|
0.4652
|
1.7312
|
1.2660
|
1.2784
|
7.8765
|
52
|
mlp_Std_Backpropagation
|
0.638
|
0.4789
|
0.5588
|
0.0799
|
0.4219
|
2.0582
|
53
|
snnR
|
snnR_none
|
0.140
|
1.9864
|
1.9864
|
0.0000
|
1.5889
|
8.8501
|
54
|
traineR
|
train.nnet_none
|
0.078
|
0.4539
|
0.5799
|
0.1260
|
0.4649
|
2.6448
|
55
|
validann
|
ann_BFGS
|
1.712
|
0.2730
|
0.4266
|
0.1536
|
0.3155
|
1.9320
|
56
|
ann_CG
|
11.228
|
0.3813
|
0.4231
|
0.0418
|
0.3165
|
1.8043
|
57
|
ann_L-BFGS-B
|
1.828
|
0.4455
|
1.5927
|
1.1472
|
1.1539
|
8.9132
|
58
|
ann_Nelder-Mead
|
2.126
|
3.1073
|
3.5453
|
0.4380
|
2.7197
|
17.3854
|
59
|
ann_SANN
|
0.172
|
3.3417
|
4.0522
|
0.7105
|
2.9633
|
19.6574
|
60
|
mFriedman
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.128
|
0.0264
|
0.0296
|
0.0032
|
0.0235
|
0.1101
|
1
|
train_ADAPTgdwm
|
0.178
|
0.0439
|
0.0450
|
0.0011
|
0.0321
|
0.1788
|
2
|
train_BATCHgd
|
1.876
|
0.0177
|
0.0816
|
0.0639
|
0.0748
|
0.1692
|
3
|
train_BATCHgdwm
|
1.882
|
0.0173
|
0.0176
|
0.0003
|
0.0138
|
0.0586
|
4
|
ANN2
|
neuralnetwork_adam
|
0.234
|
0.0183
|
0.0201
|
0.0018
|
0.0166
|
0.0579
|
5
|
neuralnetwork_rmsprop
|
0.226
|
0.0250
|
0.0314
|
0.0064
|
0.0251
|
0.0945
|
6
|
neuralnetwork_sgd
|
0.222
|
0.0178
|
0.0185
|
0.0007
|
0.0147
|
0.0603
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
9.568
|
0.0277
|
0.0323
|
0.0046
|
0.0250
|
0.1346
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
8.550
|
0.0397
|
0.0504
|
0.0107
|
0.0399
|
0.2019
|
9
|
automl_train_manual_trainwpso
|
14.836
|
0.1029
|
0.1228
|
0.0199
|
0.0976
|
0.3922
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.238
|
0.0046
|
0.0052
|
0.0006
|
0.0043
|
0.0154
|
11
|
CaDENCE
|
cadence.fit_optim
|
9.226
|
0.0160
|
0.0863
|
0.0703
|
0.0442
|
0.3640
|
12
|
cadence.fit_psoptim
|
12.056
|
0.0950
|
0.1148
|
0.0198
|
0.0739
|
0.4058
|
13
|
cadence.fit_Rprop
|
22.516
|
0.0850
|
0.1295
|
0.0445
|
0.0858
|
0.5842
|
14
|
caret
|
avNNet_none
|
0.288
|
0.0123
|
0.0197
|
0.0074
|
0.0162
|
0.0727
|
15
|
deepdive
|
deepnet_adam
|
0.768
|
0.0875
|
0.0875
|
0.0000
|
0.0764
|
0.2699
|
16
|
deepnet_gradientDescent
|
7.642
|
0.1474
|
0.1474
|
0.0000
|
0.1154
|
0.4581
|
17
|
deepnet_momentum
|
7.848
|
0.1363
|
0.1363
|
0.0000
|
0.1061
|
0.4860
|
18
|
deepnet_rmsProp
|
0.772
|
0.1287
|
0.1287
|
0.0000
|
0.0990
|
0.4133
|
19
|
deepnet
|
nn.train_BP
|
0.664
|
0.0396
|
0.0967
|
0.0571
|
0.0838
|
0.2139
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
0.1516
|
0.1734
|
0.0218
|
0.1379
|
0.5055
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.008
|
0.1677
|
0.1924
|
0.0247
|
0.1538
|
0.5716
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.092
|
0.0245
|
0.0262
|
0.0017
|
0.0181
|
0.1321
|
23
|
h2o
|
h2o.deeplearning_first-order
|
6.046
|
0.0225
|
0.0261
|
0.0036
|
0.0204
|
0.0902
|
24
|
keras
|
fit_adadelta
|
29.424
|
0.0257
|
0.0267
|
0.0010
|
0.0211
|
0.0948
|
25
|
fit_adagrad
|
14.836
|
0.0296
|
0.0842
|
0.0546
|
0.0747
|
0.2012
|
26
|
fit_adam
|
2.160
|
0.0636
|
0.0774
|
0.0138
|
0.0612
|
0.2686
|
27
|
fit_adamax
|
4.326
|
0.0326
|
0.0395
|
0.0069
|
0.0319
|
0.1140
|
28
|
fit_nadam
|
2.482
|
0.0732
|
0.0992
|
0.0260
|
0.0817
|
0.3144
|
29
|
fit_rmsprop
|
2.240
|
0.1010
|
0.1147
|
0.0137
|
0.0860
|
0.3822
|
30
|
fit_sgd
|
4.136
|
0.0365
|
0.0527
|
0.0162
|
0.0403
|
0.1922
|
31
|
MachineShop
|
fit_none
|
0.106
|
0.0085
|
0.0116
|
0.0031
|
0.0092
|
0.0360
|
32
|
minpack.lm
|
nlsLM_none
|
0.380
|
0.1269
|
0.1269
|
0.0000
|
0.1009
|
0.3714
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.308
|
0.0132
|
0.0139
|
0.0007
|
0.0110
|
0.0465
|
34
|
monmlp.fit_Nelder-Mead
|
1.084
|
0.1155
|
0.1219
|
0.0064
|
0.0960
|
0.3777
|
35
|
neuralnet
|
neuralnet_backprop
|
14.676
|
0.2348
|
0.2348
|
0.0000
|
0.1880
|
0.6346
|
36
|
neuralnet_rprop-
|
5.058
|
0.0095
|
0.0110
|
0.0015
|
0.0085
|
0.0412
|
37
|
neuralnet_rprop+
|
5.862
|
0.0102
|
0.0106
|
0.0004
|
0.0083
|
0.0356
|
38
|
neuralnet_sag
|
13.202
|
0.0806
|
0.2348
|
0.1542
|
0.1880
|
0.6346
|
39
|
neuralnet_slr
|
12.928
|
0.0690
|
0.2348
|
0.1658
|
0.1880
|
0.6346
|
40
|
nlsr
|
nlxb_none
|
0.762
|
0.0045
|
0.0061
|
0.0016
|
0.0048
|
0.0196
|
41
|
nnet
|
nnet_none
|
0.102
|
0.0091
|
0.0120
|
0.0029
|
0.0094
|
0.0404
|
42
|
qrnn
|
qrnn.fit_none
|
0.578
|
0.0105
|
0.0296
|
0.0191
|
0.0190
|
0.1330
|
43
|
radiant.model
|
nn_none
|
0.122
|
0.0084
|
0.0150
|
0.0066
|
0.0106
|
0.0664
|
44
|
rminer
|
fit_none
|
0.286
|
0.0095
|
0.0112
|
0.0017
|
0.0088
|
0.0373
|
45
|
RSNNS
|
mlp_BackpropBatch
|
6.886
|
0.0434
|
0.0851
|
0.0417
|
0.0754
|
0.2084
|
46
|
mlp_BackpropChunk
|
0.732
|
0.0541
|
0.0657
|
0.0116
|
0.0532
|
0.2284
|
47
|
mlp_BackpropMomentum
|
0.706
|
0.0558
|
0.0789
|
0.0231
|
0.0582
|
0.2590
|
48
|
mlp_BackpropWeightDecay
|
0.726
|
0.0429
|
0.0595
|
0.0166
|
0.0488
|
0.1832
|
49
|
mlp_Quickprop
|
7.502
|
0.1664
|
0.1722
|
0.0058
|
0.1384
|
0.5541
|
50
|
mlp_Rprop
|
0.706
|
0.0307
|
0.0452
|
0.0145
|
0.0374
|
0.1660
|
51
|
mlp_SCG
|
1.140
|
0.0202
|
0.0218
|
0.0016
|
0.0170
|
0.0747
|
52
|
mlp_Std_Backpropagation
|
0.694
|
0.0420
|
0.0900
|
0.0480
|
0.0761
|
0.2168
|
53
|
snnR
|
snnR_none
|
0.102
|
0.0457
|
0.0839
|
0.0382
|
0.0747
|
0.2113
|
54
|
traineR
|
train.nnet_none
|
0.094
|
0.0112
|
0.0263
|
0.0151
|
0.0209
|
0.0954
|
55
|
validann
|
ann_BFGS
|
2.500
|
0.0096
|
0.0688
|
0.0592
|
0.0500
|
0.1513
|
56
|
ann_CG
|
25.592
|
0.0163
|
0.0184
|
0.0021
|
0.0145
|
0.0580
|
57
|
ann_L-BFGS-B
|
2.702
|
0.0211
|
0.0256
|
0.0045
|
0.0208
|
0.0865
|
58
|
ann_Nelder-Mead
|
6.334
|
0.0991
|
0.1082
|
0.0091
|
0.0820
|
0.3701
|
59
|
ann_SANN
|
0.204
|
0.1414
|
0.1485
|
0.0071
|
0.1149
|
0.5629
|
60
|
mIshigami
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.222
|
0.7690
|
0.8135
|
0.0445
|
0.6083
|
2.9968
|
1
|
train_ADAPTgdwm
|
0.330
|
0.8636
|
0.9950
|
0.1314
|
0.7280
|
3.8394
|
2
|
train_BATCHgd
|
2.624
|
2.5215
|
2.5544
|
0.0329
|
2.1768
|
6.3018
|
3
|
train_BATCHgdwm
|
2.678
|
2.4805
|
2.5259
|
0.0454
|
2.1518
|
6.4536
|
4
|
ANN2
|
neuralnetwork_adam
|
1.058
|
0.7560
|
0.8062
|
0.0502
|
0.6130
|
3.5492
|
5
|
neuralnetwork_rmsprop
|
1.048
|
0.7045
|
0.8590
|
0.1545
|
0.6409
|
2.9940
|
6
|
neuralnetwork_sgd
|
1.048
|
0.7787
|
0.9097
|
0.1310
|
0.6798
|
3.8085
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
9.932
|
0.7511
|
0.7995
|
0.0484
|
0.6120
|
2.9212
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
8.882
|
1.8225
|
2.5662
|
0.7437
|
2.1749
|
6.0520
|
9
|
automl_train_manual_trainwpso
|
25.376
|
1.8381
|
2.4317
|
0.5936
|
1.9867
|
7.8872
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.204
|
0.6588
|
0.6635
|
0.0047
|
0.5100
|
2.9395
|
11
|
CaDENCE
|
cadence.fit_optim
|
14.912
|
1.0465
|
1.6993
|
0.6528
|
0.8815
|
5.3208
|
12
|
cadence.fit_psoptim
|
14.936
|
2.6775
|
2.7432
|
0.0657
|
2.3281
|
8.8488
|
13
|
cadence.fit_Rprop
|
36.926
|
1.3422
|
2.3133
|
0.9711
|
1.3927
|
8.8022
|
14
|
caret
|
avNNet_none
|
0.418
|
1.0310
|
1.6339
|
0.6029
|
1.3615
|
4.7983
|
15
|
deepdive
|
deepnet_adam
|
0.902
|
2.5913
|
2.5913
|
0.0000
|
2.0819
|
10.0604
|
16
|
deepnet_gradientDescent
|
9.062
|
3.0218
|
3.0218
|
0.0000
|
2.4940
|
10.2360
|
17
|
deepnet_momentum
|
9.220
|
2.5791
|
2.5791
|
0.0000
|
2.0107
|
8.7569
|
18
|
deepnet_rmsProp
|
0.892
|
2.6728
|
2.6728
|
0.0000
|
2.3060
|
7.1452
|
19
|
deepnet
|
nn.train_BP
|
0.770
|
1.0536
|
1.4687
|
0.4151
|
1.0190
|
6.8677
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
3.0949
|
3.2590
|
0.1641
|
2.6511
|
11.3823
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.008
|
3.2348
|
3.2840
|
0.0492
|
2.6674
|
12.0160
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.132
|
0.6342
|
0.8141
|
0.1799
|
0.5735
|
3.9523
|
23
|
h2o
|
h2o.deeplearning_first-order
|
6.462
|
0.8347
|
0.8467
|
0.0120
|
0.6295
|
3.6234
|
24
|
keras
|
fit_adadelta
|
31.676
|
2.4074
|
2.6007
|
0.1933
|
2.2281
|
6.9184
|
25
|
fit_adagrad
|
31.856
|
0.8522
|
2.5746
|
1.7224
|
2.1958
|
6.9534
|
26
|
fit_adam
|
2.796
|
0.9777
|
1.0728
|
0.0951
|
0.7886
|
4.0357
|
27
|
fit_adamax
|
5.302
|
0.8307
|
0.8615
|
0.0308
|
0.6388
|
3.6379
|
28
|
fit_nadam
|
3.264
|
1.0800
|
2.7592
|
1.6792
|
2.3587
|
8.0273
|
29
|
fit_rmsprop
|
1.924
|
2.8335
|
3.0118
|
0.1783
|
2.4550
|
9.4367
|
30
|
fit_sgd
|
2.788
|
2.7076
|
2.7302
|
0.0226
|
2.3252
|
7.5362
|
31
|
MachineShop
|
fit_none
|
0.152
|
0.6685
|
2.1956
|
1.5271
|
1.7857
|
5.3089
|
32
|
minpack.lm
|
nlsLM_none
|
0.940
|
2.5379
|
2.5379
|
0.0000
|
2.0524
|
7.6035
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.460
|
0.8185
|
0.9739
|
0.1554
|
0.7577
|
3.6164
|
34
|
monmlp.fit_Nelder-Mead
|
1.600
|
2.7368
|
2.8463
|
0.1095
|
2.3257
|
8.7509
|
35
|
neuralnet
|
neuralnet_backprop
|
23.958
|
3.6898
|
3.6898
|
0.0000
|
2.9776
|
13.1137
|
36
|
neuralnet_rprop-
|
1.954
|
0.6728
|
0.7126
|
0.0398
|
0.5316
|
2.8674
|
37
|
neuralnet_rprop+
|
4.596
|
0.5788
|
0.6650
|
0.0862
|
0.5052
|
2.7746
|
38
|
neuralnet_sag
|
25.218
|
3.6898
|
3.6898
|
0.0000
|
2.9776
|
13.1137
|
39
|
neuralnet_slr
|
24.586
|
0.6816
|
3.6898
|
3.0082
|
2.9776
|
13.1137
|
40
|
nlsr
|
nlxb_none
|
1.470
|
0.6602
|
2.2311
|
1.5709
|
1.8053
|
5.7864
|
41
|
nnet
|
nnet_none
|
0.152
|
0.5462
|
0.6959
|
0.1497
|
0.5147
|
3.0034
|
42
|
qrnn
|
qrnn.fit_none
|
1.122
|
0.7656
|
0.7907
|
0.0251
|
0.4951
|
4.0838
|
43
|
radiant.model
|
nn_none
|
0.172
|
0.4934
|
0.7868
|
0.2934
|
0.5896
|
3.1250
|
44
|
rminer
|
fit_none
|
0.446
|
0.6490
|
0.6668
|
0.0178
|
0.5016
|
3.0019
|
45
|
RSNNS
|
mlp_BackpropBatch
|
8.542
|
2.6668
|
2.6742
|
0.0074
|
2.3004
|
7.1688
|
46
|
mlp_BackpropChunk
|
0.816
|
1.3784
|
2.6226
|
1.2442
|
2.0664
|
8.9928
|
47
|
mlp_BackpropMomentum
|
0.822
|
2.6138
|
2.6595
|
0.0457
|
2.1268
|
10.0368
|
48
|
mlp_BackpropWeightDecay
|
0.854
|
1.2711
|
2.0728
|
0.8017
|
1.5275
|
7.9148
|
49
|
mlp_Quickprop
|
9.656
|
3.4245
|
3.5389
|
0.1144
|
2.8752
|
13.1137
|
50
|
mlp_Rprop
|
0.840
|
1.3146
|
2.3451
|
1.0305
|
1.8953
|
6.5010
|
51
|
mlp_SCG
|
1.456
|
0.6980
|
0.7363
|
0.0383
|
0.5439
|
3.0529
|
52
|
mlp_Std_Backpropagation
|
0.814
|
2.7659
|
2.8040
|
0.0381
|
2.1912
|
11.0805
|
53
|
snnR
|
snnR_none
|
0.430
|
0.7757
|
0.8621
|
0.0864
|
0.6030
|
3.4730
|
54
|
traineR
|
train.nnet_none
|
0.152
|
0.6846
|
0.7400
|
0.0554
|
0.5453
|
3.2851
|
55
|
validann
|
ann_BFGS
|
5.086
|
0.6342
|
0.7284
|
0.0942
|
0.5216
|
3.3533
|
56
|
ann_CG
|
58.524
|
0.6427
|
0.7212
|
0.0785
|
0.5352
|
3.3323
|
57
|
ann_L-BFGS-B
|
5.418
|
0.8502
|
1.1103
|
0.2601
|
0.8812
|
3.5016
|
58
|
ann_Nelder-Mead
|
16.114
|
2.6029
|
2.6812
|
0.0783
|
2.2886
|
7.2908
|
59
|
ann_SANN
|
0.270
|
2.9199
|
2.9986
|
0.0787
|
2.4922
|
10.0706
|
60
|
mRef153
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.038
|
3.3184
|
3.3402
|
0.0218
|
2.3679
|
13.7131
|
1
|
train_ADAPTgdwm
|
0.050
|
3.3209
|
3.5541
|
0.2332
|
2.5492
|
14.1047
|
2
|
train_BATCHgd
|
1.436
|
3.3610
|
3.5412
|
0.1802
|
2.5557
|
13.2065
|
3
|
train_BATCHgdwm
|
1.438
|
3.3094
|
3.4478
|
0.1384
|
2.4907
|
13.6270
|
4
|
ANN2
|
neuralnetwork_adam
|
0.042
|
3.4586
|
3.4861
|
0.0275
|
2.5495
|
13.6835
|
5
|
neuralnetwork_rmsprop
|
0.050
|
3.2991
|
3.3581
|
0.0590
|
2.4231
|
13.0278
|
6
|
neuralnetwork_sgd
|
0.046
|
3.5190
|
3.6375
|
0.1185
|
2.6189
|
13.4490
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
2.760
|
3.4616
|
3.7944
|
0.3328
|
2.7431
|
15.5607
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
2.792
|
3.5937
|
3.9015
|
0.3078
|
2.7453
|
14.5783
|
9
|
automl_train_manual_trainwpso
|
6.730
|
3.7011
|
3.9237
|
0.2226
|
2.8771
|
13.9777
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.020
|
3.4791
|
3.4796
|
0.0005
|
2.5068
|
13.7382
|
11
|
CaDENCE
|
cadence.fit_optim
|
3.666
|
3.2677
|
3.7049
|
0.4372
|
2.6819
|
15.6761
|
12
|
cadence.fit_psoptim
|
7.650
|
3.7348
|
3.8749
|
0.1401
|
2.7851
|
15.8749
|
13
|
cadence.fit_Rprop
|
9.946
|
3.7332
|
4.3519
|
0.6187
|
3.0787
|
16.5513
|
14
|
caret
|
avNNet_none
|
0.042
|
3.2329
|
3.3749
|
0.1420
|
2.3825
|
12.7439
|
15
|
deepdive
|
deepnet_adam
|
0.638
|
3.7605
|
3.7605
|
0.0000
|
2.6651
|
15.3279
|
16
|
deepnet_gradientDescent
|
5.964
|
4.1805
|
4.1805
|
0.0000
|
3.1843
|
14.4946
|
17
|
deepnet_momentum
|
6.090
|
3.8424
|
3.8424
|
0.0000
|
2.7879
|
15.0474
|
18
|
deepnet_rmsProp
|
0.638
|
4.0384
|
4.0384
|
0.0000
|
3.0290
|
15.7099
|
19
|
deepnet
|
nn.train_BP
|
0.206
|
3.4561
|
3.8740
|
0.4179
|
2.9819
|
15.8992
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
16.4917
|
18.7626
|
2.2709
|
15.3993
|
53.2953
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.000
|
9.4764
|
15.5973
|
6.1209
|
12.5855
|
41.9893
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.016
|
3.0740
|
3.4380
|
0.3640
|
2.4557
|
13.5954
|
23
|
h2o
|
h2o.deeplearning_first-order
|
4.408
|
3.2514
|
3.5349
|
0.2835
|
2.5500
|
14.8677
|
24
|
keras
|
fit_adadelta
|
7.872
|
3.8172
|
3.9063
|
0.0891
|
2.9028
|
14.4159
|
25
|
fit_adagrad
|
3.186
|
3.8952
|
3.9503
|
0.0551
|
2.9457
|
13.8555
|
26
|
fit_adam
|
1.430
|
3.9276
|
4.1763
|
0.2487
|
3.1624
|
14.4220
|
27
|
fit_adamax
|
2.194
|
3.7083
|
3.9479
|
0.2396
|
2.9160
|
15.5093
|
28
|
fit_nadam
|
1.434
|
4.1342
|
4.5026
|
0.3684
|
3.4678
|
16.1957
|
29
|
fit_rmsprop
|
0.996
|
4.5931
|
6.6007
|
2.0076
|
5.4044
|
17.1610
|
30
|
fit_sgd
|
2.398
|
3.7419
|
4.2292
|
0.4873
|
3.1633
|
15.5181
|
31
|
MachineShop
|
fit_none
|
0.014
|
3.1772
|
3.4737
|
0.2965
|
2.4888
|
13.8862
|
32
|
minpack.lm
|
nlsLM_none
|
0.068
|
3.4318
|
3.4318
|
0.0000
|
2.4646
|
13.2825
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.224
|
3.2360
|
3.2363
|
0.0003
|
2.2071
|
14.2052
|
34
|
monmlp.fit_Nelder-Mead
|
0.544
|
3.7710
|
4.1087
|
0.3377
|
3.0462
|
15.0511
|
35
|
neuralnet
|
neuralnet_backprop
|
0.104
|
3.8593
|
4.2101
|
0.3508
|
3.1568
|
14.3890
|
36
|
neuralnet_rprop-
|
0.010
|
3.4657
|
3.9379
|
0.4722
|
2.8577
|
15.5126
|
37
|
neuralnet_rprop+
|
0.018
|
3.4475
|
3.9946
|
0.5471
|
2.8060
|
16.5481
|
38
|
neuralnet_sag
|
0.874
|
3.3829
|
3.7868
|
0.4039
|
2.6065
|
16.3454
|
39
|
neuralnet_slr
|
0.026
|
3.8201
|
3.9883
|
0.1682
|
2.9187
|
15.3885
|
40
|
nlsr
|
nlxb_none
|
0.116
|
3.1128
|
3.4349
|
0.3221
|
2.4566
|
14.5501
|
41
|
nnet
|
nnet_none
|
0.012
|
3.0865
|
3.2253
|
0.1388
|
2.2336
|
13.9974
|
42
|
qrnn
|
qrnn.fit_none
|
0.180
|
3.2636
|
3.4540
|
0.1904
|
2.1362
|
15.2840
|
43
|
radiant.model
|
nn_none
|
0.050
|
3.1803
|
3.3593
|
0.1790
|
2.4590
|
13.9894
|
44
|
rminer
|
fit_none
|
0.048
|
3.1128
|
3.1819
|
0.0691
|
2.1870
|
13.5191
|
45
|
RSNNS
|
mlp_BackpropBatch
|
2.068
|
3.5172
|
3.6890
|
0.1718
|
2.7410
|
14.2974
|
46
|
mlp_BackpropChunk
|
0.212
|
3.5933
|
3.8022
|
0.2089
|
2.7830
|
14.4056
|
47
|
mlp_BackpropMomentum
|
0.210
|
3.4853
|
4.0633
|
0.5780
|
3.1325
|
14.3828
|
48
|
mlp_BackpropWeightDecay
|
0.246
|
3.4319
|
3.7632
|
0.3313
|
2.8900
|
14.3467
|
49
|
mlp_Quickprop
|
2.094
|
6.3871
|
8.1528
|
1.7657
|
6.2517
|
24.5588
|
50
|
mlp_Rprop
|
0.202
|
3.1838
|
3.5905
|
0.4067
|
2.5733
|
14.7429
|
51
|
mlp_SCG
|
0.326
|
3.2498
|
3.2841
|
0.0343
|
2.2854
|
14.4231
|
52
|
mlp_Std_Backpropagation
|
0.234
|
3.5707
|
3.6319
|
0.0612
|
2.6701
|
14.8068
|
53
|
snnR
|
snnR_none
|
0.020
|
4.3529
|
4.3529
|
0.0000
|
3.2423
|
15.2576
|
54
|
traineR
|
train.nnet_none
|
0.014
|
3.1128
|
3.3913
|
0.2785
|
2.4098
|
13.3731
|
55
|
validann
|
ann_BFGS
|
0.904
|
3.2262
|
3.3341
|
0.1079
|
2.3356
|
14.3369
|
56
|
ann_CG
|
30.460
|
3.1175
|
3.2299
|
0.1124
|
2.1865
|
13.2863
|
57
|
ann_L-BFGS-B
|
1.038
|
3.2380
|
3.5703
|
0.3323
|
2.6699
|
14.7850
|
58
|
ann_Nelder-Mead
|
22.952
|
3.9234
|
4.1325
|
0.2091
|
3.1924
|
14.4615
|
59
|
ann_SANN
|
0.140
|
5.7592
|
7.3556
|
1.5964
|
5.9016
|
19.8265
|
60
|
Packages Tested per Univariate Dataset
uDmod1
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.036
|
0.3082
|
0.3271
|
0.0189
|
0.2829
|
0.7263
|
1
|
train_ADAPTgdwm
|
0.054
|
0.2197
|
0.2765
|
0.0568
|
0.2204
|
0.6575
|
2
|
train_BATCHgd
|
1.780
|
0.2023
|
0.2922
|
0.0899
|
0.2393
|
0.6802
|
3
|
train_BATCHgdwm
|
1.804
|
0.3265
|
0.3274
|
0.0009
|
0.2853
|
0.7289
|
4
|
ANN2
|
neuralnetwork_adam
|
0.012
|
0.2198
|
0.2274
|
0.0076
|
0.1806
|
0.5242
|
5
|
neuralnetwork_rmsprop
|
0.016
|
0.2345
|
0.2495
|
0.0150
|
0.1926
|
0.6040
|
6
|
neuralnetwork_sgd
|
0.014
|
0.2581
|
0.3342
|
0.0761
|
0.2899
|
0.6824
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
1.262
|
0.0596
|
0.1157
|
0.0561
|
0.0741
|
0.5060
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
1.128
|
0.1052
|
0.1595
|
0.0543
|
0.1323
|
0.3299
|
9
|
automl_train_manual_trainwpso
|
6.964
|
0.2424
|
0.2517
|
0.0093
|
0.1929
|
0.6461
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.010
|
0.0451
|
0.5884
|
0.5433
|
0.5069
|
1.0104
|
11
|
CaDENCE
|
cadence.fit_optim
|
2.442
|
0.0564
|
0.2112
|
0.1548
|
0.1061
|
0.6888
|
12
|
cadence.fit_psoptim
|
5.378
|
0.3096
|
0.3190
|
0.0094
|
0.2672
|
0.7427
|
13
|
cadence.fit_Rprop
|
6.654
|
0.2005
|
0.4116
|
0.2111
|
0.3162
|
0.8665
|
14
|
caret
|
avNNet_none
|
0.030
|
0.0535
|
0.0948
|
0.0413
|
0.0602
|
0.3176
|
15
|
deepdive
|
deepnet_adam
|
0.568
|
0.1178
|
0.1178
|
0.0000
|
0.0797
|
0.4868
|
16
|
deepnet_gradientDescent
|
5.340
|
0.3353
|
0.3353
|
0.0000
|
0.2912
|
0.7067
|
17
|
deepnet_momentum
|
5.524
|
0.3320
|
0.3320
|
0.0000
|
0.2891
|
0.7441
|
18
|
deepnet_rmsProp
|
0.584
|
0.1728
|
0.1728
|
0.0000
|
0.1257
|
0.4478
|
19
|
deepnet
|
nn.train_BP
|
0.094
|
0.0582
|
0.1173
|
0.0591
|
0.0845
|
0.3896
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
0.3320
|
0.3623
|
0.0303
|
0.3038
|
0.8727
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.000
|
0.3003
|
0.3082
|
0.0079
|
0.2529
|
0.7867
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.004
|
0.0733
|
0.1033
|
0.0300
|
0.0759
|
0.4193
|
23
|
h2o
|
h2o.deeplearning_first-order
|
3.346
|
0.0480
|
0.0494
|
0.0014
|
0.0402
|
0.1185
|
24
|
keras
|
fit_adadelta
|
23.966
|
0.2314
|
0.2333
|
0.0019
|
0.1843
|
0.5698
|
25
|
fit_adagrad
|
8.322
|
0.2252
|
0.3529
|
0.1277
|
0.3037
|
0.8099
|
26
|
fit_adam
|
2.576
|
0.1376
|
0.1811
|
0.0435
|
0.1461
|
0.4721
|
27
|
fit_adamax
|
4.566
|
0.0883
|
0.2240
|
0.1357
|
0.1782
|
0.5854
|
28
|
fit_nadam
|
2.242
|
0.1786
|
0.2607
|
0.0821
|
0.2055
|
0.6971
|
29
|
fit_rmsprop
|
1.326
|
0.2375
|
0.3800
|
0.1425
|
0.2964
|
0.8503
|
30
|
fit_sgd
|
2.644
|
0.2044
|
0.3548
|
0.1504
|
0.2992
|
0.8224
|
31
|
MachineShop
|
fit_none
|
0.012
|
0.0442
|
0.0456
|
0.0014
|
0.0365
|
0.1181
|
32
|
minpack.lm
|
nlsLM_none
|
0.038
|
0.0445
|
0.0445
|
0.0000
|
0.0362
|
0.1153
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.208
|
0.0919
|
0.0983
|
0.0064
|
0.0750
|
0.3693
|
34
|
monmlp.fit_Nelder-Mead
|
0.428
|
0.1381
|
0.2639
|
0.1258
|
0.2153
|
0.6177
|
35
|
neuralnet
|
neuralnet_backprop
|
0.490
|
0.1521
|
0.1699
|
0.0178
|
0.1280
|
0.5924
|
36
|
neuralnet_rprop-
|
0.030
|
0.1634
|
0.1750
|
0.0116
|
0.1370
|
0.5212
|
37
|
neuralnet_rprop+
|
0.042
|
0.1086
|
0.1639
|
0.0553
|
0.1319
|
0.5153
|
38
|
neuralnet_sag
|
1.430
|
0.0583
|
0.1315
|
0.0732
|
0.1061
|
0.3669
|
39
|
neuralnet_slr
|
0.100
|
0.0839
|
0.1213
|
0.0374
|
0.0922
|
0.3196
|
40
|
nlsr
|
nlxb_none
|
0.088
|
0.0433
|
0.0433
|
0.0000
|
0.0349
|
0.1063
|
41
|
nnet
|
nnet_none
|
0.008
|
0.0437
|
0.0865
|
0.0428
|
0.0636
|
0.3435
|
42
|
qrnn
|
qrnn.fit_none
|
0.230
|
0.1162
|
0.1349
|
0.0187
|
0.0830
|
0.6014
|
43
|
radiant.model
|
nn_none
|
0.026
|
0.0800
|
0.1088
|
0.0288
|
0.0817
|
0.3346
|
44
|
rminer
|
fit_none
|
0.030
|
0.0449
|
0.0495
|
0.0046
|
0.0418
|
0.1258
|
45
|
RSNNS
|
mlp_BackpropBatch
|
0.874
|
0.2568
|
0.3344
|
0.0776
|
0.2870
|
0.7691
|
46
|
mlp_BackpropChunk
|
0.140
|
0.1298
|
0.1448
|
0.0150
|
0.1073
|
0.5245
|
47
|
mlp_BackpropMomentum
|
0.088
|
0.1445
|
0.1647
|
0.0202
|
0.1252
|
0.5800
|
48
|
mlp_BackpropWeightDecay
|
0.090
|
0.1314
|
0.1656
|
0.0342
|
0.1218
|
0.5395
|
49
|
mlp_Quickprop
|
0.938
|
0.5775
|
0.5884
|
0.0109
|
0.5068
|
1.0104
|
50
|
mlp_Rprop
|
0.090
|
0.1232
|
0.1401
|
0.0169
|
0.1048
|
0.4453
|
51
|
mlp_SCG
|
0.140
|
0.0970
|
0.1118
|
0.0148
|
0.0916
|
0.4280
|
52
|
mlp_Std_Backpropagation
|
0.094
|
0.1215
|
0.2226
|
0.1011
|
0.1736
|
0.5618
|
53
|
snnR
|
snnR_none
|
0.040
|
0.2927
|
0.2927
|
0.0000
|
0.2512
|
0.6561
|
54
|
traineR
|
train.nnet_none
|
0.004
|
0.0410
|
0.0470
|
0.0060
|
0.0393
|
0.1293
|
55
|
validann
|
ann_BFGS
|
0.790
|
0.0435
|
0.0725
|
0.0290
|
0.0540
|
0.1810
|
56
|
ann_CG
|
29.066
|
0.0506
|
0.0679
|
0.0173
|
0.0544
|
0.1577
|
57
|
ann_L-BFGS-B
|
0.880
|
0.0489
|
0.1090
|
0.0601
|
0.0759
|
0.4093
|
58
|
ann_Nelder-Mead
|
28.208
|
0.1034
|
0.1810
|
0.0776
|
0.1538
|
0.4017
|
59
|
ann_SANN
|
0.128
|
0.2296
|
0.3046
|
0.0750
|
0.2441
|
0.6614
|
60
|
uDmod2
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.022
|
0.2579
|
0.2632
|
0.0053
|
0.2333
|
0.5307
|
1
|
train_ADAPTgdwm
|
0.034
|
0.1145
|
0.1924
|
0.0779
|
0.1573
|
0.4195
|
2
|
train_BATCHgd
|
1.650
|
0.2228
|
0.2644
|
0.0416
|
0.2347
|
0.4989
|
3
|
train_BATCHgdwm
|
1.650
|
0.1585
|
0.2621
|
0.1036
|
0.2341
|
0.4898
|
4
|
ANN2
|
neuralnetwork_adam
|
0.014
|
0.1702
|
0.2126
|
0.0424
|
0.1747
|
0.4630
|
5
|
neuralnetwork_rmsprop
|
0.012
|
0.1831
|
0.2585
|
0.0754
|
0.2227
|
0.5514
|
6
|
neuralnetwork_sgd
|
0.012
|
0.2518
|
0.2732
|
0.0214
|
0.2401
|
0.5272
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
1.260
|
0.0511
|
0.0867
|
0.0356
|
0.0707
|
0.1976
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
1.104
|
0.1245
|
0.2296
|
0.1051
|
0.1669
|
0.5276
|
9
|
automl_train_manual_trainwpso
|
10.432
|
0.2032
|
0.2573
|
0.0541
|
0.2232
|
0.5240
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.020
|
0.0435
|
0.0673
|
0.0238
|
0.0522
|
0.1838
|
11
|
CaDENCE
|
cadence.fit_optim
|
2.310
|
0.0688
|
0.0805
|
0.0117
|
0.0582
|
0.2385
|
12
|
cadence.fit_psoptim
|
5.052
|
0.2114
|
0.3238
|
0.1124
|
0.2711
|
0.6808
|
13
|
cadence.fit_Rprop
|
5.784
|
0.1820
|
0.2615
|
0.0795
|
0.2061
|
0.6887
|
14
|
caret
|
avNNet_none
|
0.024
|
0.0512
|
0.0558
|
0.0046
|
0.0454
|
0.1561
|
15
|
deepdive
|
deepnet_adam
|
0.572
|
0.3189
|
0.3189
|
0.0000
|
0.2113
|
0.7209
|
16
|
deepnet_gradientDescent
|
5.326
|
0.2699
|
0.2699
|
0.0000
|
0.2369
|
0.5509
|
17
|
deepnet_momentum
|
5.460
|
0.2656
|
0.2656
|
0.0000
|
0.2355
|
0.5269
|
18
|
deepnet_rmsProp
|
0.558
|
0.2252
|
0.2252
|
0.0000
|
0.1580
|
0.5513
|
19
|
deepnet
|
nn.train_BP
|
0.092
|
0.0563
|
0.0608
|
0.0045
|
0.0490
|
0.1446
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
0.2589
|
0.2648
|
0.0059
|
0.2308
|
0.5419
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.000
|
0.2613
|
0.2735
|
0.0122
|
0.2358
|
0.5366
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.006
|
0.0618
|
0.0638
|
0.0020
|
0.0504
|
0.1716
|
23
|
h2o
|
h2o.deeplearning_first-order
|
3.342
|
0.0474
|
0.0482
|
0.0008
|
0.0394
|
0.1126
|
24
|
keras
|
fit_adadelta
|
26.680
|
0.1746
|
0.1792
|
0.0046
|
0.1379
|
0.4116
|
25
|
fit_adagrad
|
13.616
|
0.1597
|
0.1792
|
0.0195
|
0.1379
|
0.4143
|
26
|
fit_adam
|
2.338
|
0.0963
|
0.1767
|
0.0804
|
0.1397
|
0.4913
|
27
|
fit_adamax
|
3.696
|
0.1082
|
0.1728
|
0.0646
|
0.1393
|
0.3885
|
28
|
fit_nadam
|
2.460
|
0.1201
|
0.1884
|
0.0683
|
0.1486
|
0.5214
|
29
|
fit_rmsprop
|
1.744
|
0.1629
|
0.2166
|
0.0537
|
0.1697
|
0.5175
|
30
|
fit_sgd
|
1.868
|
0.2431
|
0.3056
|
0.0625
|
0.2606
|
0.6923
|
31
|
MachineShop
|
fit_none
|
0.012
|
0.0406
|
0.0494
|
0.0088
|
0.0374
|
0.1288
|
32
|
minpack.lm
|
nlsLM_none
|
0.024
|
0.0427
|
0.0427
|
0.0000
|
0.0333
|
0.1058
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.210
|
0.0522
|
0.0796
|
0.0274
|
0.0625
|
0.2280
|
34
|
monmlp.fit_Nelder-Mead
|
0.370
|
0.1342
|
0.1780
|
0.0438
|
0.1371
|
0.4534
|
35
|
neuralnet
|
neuralnet_backprop
|
0.372
|
0.1091
|
0.1355
|
0.0264
|
0.1084
|
0.3411
|
36
|
neuralnet_rprop-
|
0.062
|
0.0955
|
0.1186
|
0.0231
|
0.0920
|
0.2812
|
37
|
neuralnet_rprop+
|
0.036
|
0.1077
|
0.1207
|
0.0130
|
0.0926
|
0.2648
|
38
|
neuralnet_sag
|
0.950
|
0.0811
|
0.1160
|
0.0349
|
0.0943
|
0.2960
|
39
|
neuralnet_slr
|
0.092
|
0.0840
|
0.1039
|
0.0199
|
0.0866
|
0.2554
|
40
|
nlsr
|
nlxb_none
|
0.036
|
0.0427
|
0.0427
|
0.0000
|
0.0333
|
0.1058
|
41
|
nnet
|
nnet_none
|
0.008
|
0.0602
|
0.0615
|
0.0013
|
0.0489
|
0.1408
|
42
|
qrnn
|
qrnn.fit_none
|
0.214
|
0.0511
|
0.0821
|
0.0310
|
0.0598
|
0.2411
|
43
|
radiant.model
|
nn_none
|
0.022
|
0.0647
|
0.0771
|
0.0124
|
0.0602
|
0.2202
|
44
|
rminer
|
fit_none
|
0.016
|
0.0405
|
0.0579
|
0.0174
|
0.0479
|
0.1065
|
45
|
RSNNS
|
mlp_BackpropBatch
|
0.870
|
0.2601
|
0.2736
|
0.0135
|
0.2371
|
0.6099
|
46
|
mlp_BackpropChunk
|
0.090
|
0.0829
|
0.0892
|
0.0063
|
0.0732
|
0.2035
|
47
|
mlp_BackpropMomentum
|
0.092
|
0.0752
|
0.0964
|
0.0212
|
0.0786
|
0.2134
|
48
|
mlp_BackpropWeightDecay
|
0.094
|
0.0799
|
0.0888
|
0.0089
|
0.0704
|
0.2063
|
49
|
mlp_Quickprop
|
0.906
|
0.2570
|
0.4804
|
0.2234
|
0.4177
|
1.0187
|
50
|
mlp_Rprop
|
0.090
|
0.0447
|
0.0959
|
0.0512
|
0.0717
|
0.2622
|
51
|
mlp_SCG
|
0.132
|
0.0555
|
0.0788
|
0.0233
|
0.0618
|
0.2070
|
52
|
mlp_Std_Backpropagation
|
0.086
|
0.0788
|
0.1292
|
0.0504
|
0.0999
|
0.3342
|
53
|
snnR
|
snnR_none
|
0.020
|
0.2585
|
0.2984
|
0.0399
|
0.2556
|
0.6651
|
54
|
traineR
|
train.nnet_none
|
0.016
|
0.0505
|
0.0649
|
0.0144
|
0.0529
|
0.1392
|
55
|
validann
|
ann_BFGS
|
0.676
|
0.0405
|
0.0437
|
0.0032
|
0.0342
|
0.1131
|
56
|
ann_CG
|
31.616
|
0.0536
|
0.0610
|
0.0074
|
0.0476
|
0.1415
|
57
|
ann_L-BFGS-B
|
0.784
|
0.0691
|
0.0778
|
0.0087
|
0.0617
|
0.2019
|
58
|
ann_Nelder-Mead
|
30.686
|
0.0633
|
0.1987
|
0.1354
|
0.1673
|
0.4305
|
59
|
ann_SANN
|
0.148
|
0.2274
|
0.2546
|
0.0272
|
0.1990
|
0.5099
|
60
|
uDreyfus1
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.020
|
0.3308
|
0.3475
|
0.0167
|
0.2718
|
0.7716
|
1
|
train_ADAPTgdwm
|
0.030
|
0.1804
|
0.2112
|
0.0308
|
0.1476
|
0.4856
|
2
|
train_BATCHgd
|
1.382
|
0.3160
|
0.3346
|
0.0186
|
0.2740
|
0.7087
|
3
|
train_BATCHgdwm
|
1.376
|
0.3346
|
0.3370
|
0.0024
|
0.2785
|
0.7142
|
4
|
ANN2
|
neuralnetwork_adam
|
0.006
|
0.2762
|
0.3201
|
0.0439
|
0.2542
|
0.7372
|
5
|
neuralnetwork_rmsprop
|
0.006
|
0.2467
|
0.3428
|
0.0961
|
0.2715
|
0.7616
|
6
|
neuralnetwork_sgd
|
0.008
|
0.3493
|
0.3546
|
0.0053
|
0.2689
|
0.8481
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
0.906
|
0.0087
|
0.0725
|
0.0638
|
0.0481
|
0.2070
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
1.114
|
0.0479
|
0.0727
|
0.0248
|
0.0498
|
0.2335
|
9
|
automl_train_manual_trainwpso
|
5.358
|
0.1052
|
0.1154
|
0.0102
|
0.0854
|
0.3281
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.000
|
0.0026
|
0.0034
|
0.0008
|
0.0029
|
0.0115
|
11
|
CaDENCE
|
cadence.fit_optim
|
1.030
|
0.0032
|
0.6701
|
0.6669
|
0.3727
|
1.9004
|
12
|
cadence.fit_psoptim
|
4.512
|
0.4218
|
0.5720
|
0.1502
|
0.3001
|
1.6829
|
13
|
cadence.fit_Rprop
|
3.762
|
0.3995
|
1.1290
|
0.7295
|
0.8219
|
2.2557
|
14
|
caret
|
avNNet_none
|
0.018
|
0.0262
|
0.0359
|
0.0097
|
0.0282
|
0.1115
|
15
|
deepdive
|
deepnet_adam
|
0.564
|
0.0304
|
0.0304
|
0.0000
|
0.0265
|
0.0644
|
16
|
deepnet_gradientDescent
|
5.162
|
0.3429
|
0.3429
|
0.0000
|
0.2801
|
0.7346
|
17
|
deepnet_momentum
|
5.460
|
0.3429
|
0.3429
|
0.0000
|
0.2801
|
0.7341
|
18
|
deepnet_rmsProp
|
0.566
|
0.1184
|
0.1184
|
0.0000
|
0.0878
|
0.3446
|
19
|
deepnet
|
nn.train_BP
|
0.084
|
0.0139
|
0.0704
|
0.0565
|
0.0451
|
0.2219
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
0.3407
|
0.4066
|
0.0659
|
0.2973
|
1.0342
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.000
|
0.3987
|
0.4505
|
0.0518
|
0.3027
|
1.1845
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.002
|
0.0922
|
0.1151
|
0.0229
|
0.0834
|
0.3335
|
23
|
h2o
|
h2o.deeplearning_first-order
|
3.334
|
0.0131
|
0.0146
|
0.0015
|
0.0112
|
0.0432
|
24
|
keras
|
fit_adadelta
|
10.958
|
0.2178
|
0.3498
|
0.1320
|
0.2655
|
0.8040
|
25
|
fit_adagrad
|
5.918
|
0.1630
|
0.3528
|
0.1898
|
0.2697
|
0.8045
|
26
|
fit_adam
|
2.808
|
0.0706
|
0.0897
|
0.0191
|
0.0690
|
0.2075
|
27
|
fit_adamax
|
5.074
|
0.0365
|
0.0487
|
0.0122
|
0.0404
|
0.1489
|
28
|
fit_nadam
|
2.082
|
0.0648
|
0.1550
|
0.0902
|
0.1179
|
0.3970
|
29
|
fit_rmsprop
|
0.942
|
0.3101
|
0.3622
|
0.0521
|
0.2798
|
0.9265
|
30
|
fit_sgd
|
2.312
|
0.3373
|
0.3450
|
0.0077
|
0.2698
|
0.7744
|
31
|
MachineShop
|
fit_none
|
0.010
|
0.0023
|
0.0034
|
0.0011
|
0.0028
|
0.0102
|
32
|
minpack.lm
|
nlsLM_none
|
0.000
|
0.0000
|
0.0000
|
0.0000
|
0.0000
|
0.0001
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.190
|
0.0323
|
0.0541
|
0.0218
|
0.0434
|
0.1524
|
34
|
monmlp.fit_Nelder-Mead
|
0.270
|
0.1425
|
0.2017
|
0.0592
|
0.1653
|
0.4572
|
35
|
neuralnet
|
neuralnet_backprop
|
0.040
|
0.3201
|
0.3503
|
0.0302
|
0.2743
|
0.7831
|
36
|
neuralnet_rprop-
|
0.008
|
0.1014
|
0.2856
|
0.1842
|
0.2227
|
0.7157
|
37
|
neuralnet_rprop+
|
0.004
|
0.2119
|
0.3475
|
0.1356
|
0.2662
|
0.7910
|
38
|
neuralnet_sag
|
0.048
|
0.1963
|
0.3371
|
0.1408
|
0.2652
|
0.7510
|
39
|
neuralnet_slr
|
0.012
|
0.2981
|
0.3450
|
0.0469
|
0.2730
|
0.7821
|
40
|
nlsr
|
nlxb_none
|
0.014
|
0.0000
|
0.0000
|
0.0000
|
0.0000
|
0.0001
|
41
|
nnet
|
nnet_none
|
0.004
|
0.0026
|
0.0716
|
0.0690
|
0.0449
|
0.2254
|
42
|
qrnn
|
qrnn.fit_none
|
0.128
|
0.2781
|
0.2841
|
0.0060
|
0.1815
|
0.9095
|
43
|
radiant.model
|
nn_none
|
0.022
|
0.0121
|
0.0682
|
0.0561
|
0.0555
|
0.1546
|
44
|
rminer
|
fit_none
|
0.012
|
0.0020
|
0.0023
|
0.0003
|
0.0018
|
0.0057
|
45
|
RSNNS
|
mlp_BackpropBatch
|
0.812
|
0.3120
|
0.3387
|
0.0267
|
0.2647
|
0.7662
|
46
|
mlp_BackpropChunk
|
0.088
|
0.0838
|
0.1275
|
0.0437
|
0.0822
|
0.3313
|
47
|
mlp_BackpropMomentum
|
0.080
|
0.0719
|
0.0795
|
0.0076
|
0.0606
|
0.2070
|
48
|
mlp_BackpropWeightDecay
|
0.090
|
0.0797
|
0.0849
|
0.0052
|
0.0657
|
0.2541
|
49
|
mlp_Quickprop
|
0.822
|
0.2177
|
0.2408
|
0.0231
|
0.2084
|
0.5154
|
50
|
mlp_Rprop
|
0.080
|
0.0617
|
0.0689
|
0.0072
|
0.0484
|
0.2211
|
51
|
mlp_SCG
|
0.122
|
0.0851
|
0.1018
|
0.0167
|
0.0848
|
0.2408
|
52
|
mlp_Std_Backpropagation
|
0.078
|
0.1127
|
0.1190
|
0.0063
|
0.1000
|
0.2547
|
53
|
snnR
|
snnR_none
|
0.006
|
0.3691
|
0.3691
|
0.0000
|
0.2756
|
0.8531
|
54
|
traineR
|
train.nnet_none
|
0.000
|
0.0019
|
0.0022
|
0.0003
|
0.0019
|
0.0076
|
55
|
validann
|
ann_BFGS
|
0.350
|
0.0022
|
0.0023
|
0.0001
|
0.0019
|
0.0070
|
56
|
ann_CG
|
25.106
|
0.0035
|
0.0076
|
0.0041
|
0.0061
|
0.0205
|
57
|
ann_L-BFGS-B
|
0.512
|
0.0038
|
0.0084
|
0.0046
|
0.0066
|
0.0207
|
58
|
ann_Nelder-Mead
|
17.878
|
0.0833
|
0.1951
|
0.1118
|
0.1633
|
0.3615
|
59
|
ann_SANN
|
0.142
|
0.2692
|
0.3271
|
0.0579
|
0.2707
|
0.6914
|
60
|
uDreyfus2
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.020
|
0.3555
|
0.3612
|
0.0057
|
0.2824
|
0.9010
|
1
|
train_ADAPTgdwm
|
0.030
|
0.1675
|
0.2519
|
0.0844
|
0.1964
|
0.6689
|
2
|
train_BATCHgd
|
1.378
|
0.1778
|
0.2708
|
0.0930
|
0.2148
|
0.6304
|
3
|
train_BATCHgdwm
|
1.380
|
0.2097
|
0.3405
|
0.1308
|
0.2704
|
0.8640
|
4
|
ANN2
|
neuralnetwork_adam
|
0.010
|
0.3222
|
0.3836
|
0.0614
|
0.2918
|
0.8958
|
5
|
neuralnetwork_rmsprop
|
0.014
|
0.2338
|
0.2845
|
0.0507
|
0.2093
|
0.8061
|
6
|
neuralnetwork_sgd
|
0.008
|
0.3581
|
0.3717
|
0.0136
|
0.2806
|
0.9555
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
1.244
|
0.0933
|
0.1579
|
0.0646
|
0.1212
|
0.4579
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
1.096
|
0.1179
|
0.1615
|
0.0436
|
0.1223
|
0.4550
|
9
|
automl_train_manual_trainwpso
|
5.164
|
0.1180
|
0.1616
|
0.0436
|
0.1233
|
0.4403
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.000
|
0.0913
|
0.0913
|
0.0000
|
0.0730
|
0.2241
|
11
|
CaDENCE
|
cadence.fit_optim
|
1.046
|
0.0924
|
0.3856
|
0.2932
|
0.2560
|
1.1222
|
12
|
cadence.fit_psoptim
|
4.500
|
0.3210
|
0.3814
|
0.0604
|
0.2825
|
1.0638
|
13
|
cadence.fit_Rprop
|
4.272
|
0.1684
|
0.2586
|
0.0902
|
0.1963
|
0.7853
|
14
|
caret
|
avNNet_none
|
0.022
|
0.0926
|
0.1039
|
0.0113
|
0.0811
|
0.2375
|
15
|
deepdive
|
deepnet_adam
|
0.566
|
0.1149
|
0.1149
|
0.0000
|
0.0907
|
0.2749
|
16
|
deepnet_gradientDescent
|
5.222
|
0.3570
|
0.3570
|
0.0000
|
0.2905
|
0.8478
|
17
|
deepnet_momentum
|
5.390
|
0.3570
|
0.3570
|
0.0000
|
0.2907
|
0.8468
|
18
|
deepnet_rmsProp
|
0.552
|
0.1625
|
0.1625
|
0.0000
|
0.1232
|
0.4839
|
19
|
deepnet
|
nn.train_BP
|
0.080
|
0.0928
|
0.1049
|
0.0121
|
0.0824
|
0.2577
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
0.4534
|
0.6226
|
0.1692
|
0.5077
|
1.4031
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.002
|
0.4554
|
0.5844
|
0.1290
|
0.4293
|
1.3727
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.008
|
0.1196
|
0.1272
|
0.0076
|
0.1022
|
0.3102
|
23
|
h2o
|
h2o.deeplearning_first-order
|
3.356
|
0.0926
|
0.0933
|
0.0007
|
0.0740
|
0.2242
|
24
|
keras
|
fit_adadelta
|
7.248
|
0.3605
|
0.3726
|
0.0121
|
0.2763
|
0.9935
|
25
|
fit_adagrad
|
14.992
|
0.1847
|
0.1979
|
0.0132
|
0.1420
|
0.6153
|
26
|
fit_adam
|
2.354
|
0.1105
|
0.1345
|
0.0240
|
0.1062
|
0.3156
|
27
|
fit_adamax
|
4.740
|
0.1118
|
0.1154
|
0.0036
|
0.0935
|
0.3048
|
28
|
fit_nadam
|
2.366
|
0.1341
|
0.1957
|
0.0616
|
0.1530
|
0.5132
|
29
|
fit_rmsprop
|
1.096
|
0.2221
|
0.3598
|
0.1377
|
0.2805
|
0.8788
|
30
|
fit_sgd
|
2.546
|
0.3523
|
0.3548
|
0.0025
|
0.2760
|
0.9224
|
31
|
MachineShop
|
fit_none
|
0.012
|
0.0906
|
0.1415
|
0.0509
|
0.1045
|
0.4507
|
32
|
minpack.lm
|
nlsLM_none
|
0.022
|
0.0906
|
0.0906
|
0.0000
|
0.0723
|
0.2197
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.210
|
0.0917
|
0.0951
|
0.0034
|
0.0746
|
0.2363
|
34
|
monmlp.fit_Nelder-Mead
|
0.248
|
0.1762
|
0.2448
|
0.0686
|
0.1940
|
0.6268
|
35
|
neuralnet
|
neuralnet_backprop
|
0.042
|
0.3205
|
0.3655
|
0.0450
|
0.2776
|
0.9575
|
36
|
neuralnet_rprop-
|
0.014
|
0.1632
|
0.3537
|
0.1905
|
0.2756
|
0.9038
|
37
|
neuralnet_rprop+
|
0.008
|
0.2846
|
0.3562
|
0.0716
|
0.2762
|
0.8861
|
38
|
neuralnet_sag
|
0.098
|
0.1663
|
0.2521
|
0.0858
|
0.1997
|
0.7213
|
39
|
neuralnet_slr
|
0.012
|
0.3374
|
0.3435
|
0.0061
|
0.2714
|
0.9130
|
40
|
nlsr
|
nlxb_none
|
0.064
|
0.0906
|
0.0906
|
0.0000
|
0.0723
|
0.2197
|
41
|
nnet
|
nnet_none
|
0.000
|
0.0906
|
0.0906
|
0.0000
|
0.0725
|
0.2202
|
42
|
qrnn
|
qrnn.fit_none
|
0.170
|
0.1601
|
0.2693
|
0.1092
|
0.1983
|
0.6891
|
43
|
radiant.model
|
nn_none
|
0.026
|
0.0907
|
0.0917
|
0.0010
|
0.0732
|
0.2275
|
44
|
rminer
|
fit_none
|
0.018
|
0.0906
|
0.0906
|
0.0000
|
0.0724
|
0.2202
|
45
|
RSNNS
|
mlp_BackpropBatch
|
0.820
|
0.3063
|
0.3491
|
0.0428
|
0.2736
|
0.8922
|
46
|
mlp_BackpropChunk
|
0.084
|
0.1199
|
0.1689
|
0.0490
|
0.1245
|
0.5094
|
47
|
mlp_BackpropMomentum
|
0.086
|
0.1209
|
0.1297
|
0.0088
|
0.1042
|
0.3327
|
48
|
mlp_BackpropWeightDecay
|
0.082
|
0.1186
|
0.1214
|
0.0028
|
0.0950
|
0.2751
|
49
|
mlp_Quickprop
|
0.826
|
0.2122
|
0.2993
|
0.0871
|
0.2338
|
0.8131
|
50
|
mlp_Rprop
|
0.082
|
0.1145
|
0.1252
|
0.0107
|
0.1009
|
0.3338
|
51
|
mlp_SCG
|
0.114
|
0.1238
|
0.2542
|
0.1304
|
0.1876
|
0.7205
|
52
|
mlp_Std_Backpropagation
|
0.080
|
0.1298
|
0.1325
|
0.0027
|
0.1025
|
0.3352
|
53
|
snnR
|
snnR_none
|
0.012
|
0.3837
|
0.3837
|
0.0000
|
0.2773
|
1.0352
|
54
|
traineR
|
train.nnet_none
|
0.004
|
0.0906
|
0.1123
|
0.0217
|
0.0901
|
0.2736
|
55
|
validann
|
ann_BFGS
|
0.442
|
0.0906
|
0.0906
|
0.0000
|
0.0724
|
0.2200
|
56
|
ann_CG
|
26.808
|
0.0910
|
0.0913
|
0.0003
|
0.0730
|
0.2244
|
57
|
ann_L-BFGS-B
|
0.504
|
0.0907
|
0.1123
|
0.0216
|
0.0897
|
0.2733
|
58
|
ann_Nelder-Mead
|
16.024
|
0.1300
|
0.1604
|
0.0304
|
0.1224
|
0.4798
|
59
|
ann_SANN
|
0.152
|
0.2712
|
0.2972
|
0.0260
|
0.2354
|
0.7465
|
60
|
uGauss1
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.050
|
12.5180
|
28.6849
|
16.1669
|
23.0898
|
63.3445
|
1
|
train_ADAPTgdwm
|
0.084
|
14.8661
|
28.9286
|
14.0625
|
15.0901
|
75.7312
|
2
|
train_BATCHgd
|
1.694
|
12.0830
|
12.6864
|
0.6034
|
10.8841
|
27.5733
|
3
|
train_BATCHgdwm
|
1.724
|
12.0720
|
12.5131
|
0.4411
|
10.6879
|
26.8772
|
4
|
ANN2
|
neuralnetwork_adam
|
0.080
|
2.7029
|
9.5851
|
6.8822
|
7.9902
|
22.2599
|
5
|
neuralnetwork_rmsprop
|
0.076
|
5.5210
|
8.5210
|
3.0000
|
7.0532
|
20.1249
|
6
|
neuralnetwork_sgd
|
0.080
|
11.0994
|
11.9128
|
0.8134
|
10.0259
|
25.8843
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
4.982
|
4.4523
|
4.9986
|
0.5463
|
3.8522
|
18.0133
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
4.482
|
4.9906
|
5.2304
|
0.2398
|
4.1778
|
17.4681
|
9
|
automl_train_manual_trainwpso
|
8.586
|
9.8111
|
13.4226
|
3.6115
|
9.5187
|
39.8505
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.042
|
2.2434
|
2.4366
|
0.1932
|
1.8814
|
8.7824
|
11
|
CaDENCE
|
cadence.fit_optim
|
2.906
|
2.3392
|
2.4124
|
0.0732
|
1.9079
|
7.6898
|
12
|
cadence.fit_psoptim
|
6.144
|
25.3213
|
29.0788
|
3.7575
|
22.9938
|
64.9998
|
13
|
cadence.fit_Rprop
|
9.070
|
17.7666
|
25.2780
|
7.5114
|
17.3441
|
56.1202
|
14
|
caret
|
avNNet_none
|
0.108
|
2.3241
|
2.6365
|
0.3124
|
2.0948
|
7.8111
|
15
|
deepdive
|
deepnet_adam
|
0.634
|
20.5179
|
20.5179
|
0.0000
|
16.5474
|
47.6354
|
16
|
deepnet_gradientDescent
|
6.084
|
23.6597
|
23.6597
|
0.0000
|
19.1988
|
57.7174
|
17
|
deepnet_momentum
|
6.236
|
16.2557
|
16.2557
|
0.0000
|
13.0878
|
49.2409
|
18
|
deepnet_rmsProp
|
0.658
|
23.5833
|
23.5833
|
0.0000
|
19.1113
|
57.9958
|
19
|
deepnet
|
nn.train_BP
|
0.304
|
3.4191
|
4.2839
|
0.8648
|
3.5260
|
11.8002
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
17.0430
|
18.7830
|
1.7400
|
15.4373
|
50.7429
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.010
|
37.3986
|
490.0535
|
452.6549
|
426.0995
|
1516.4503
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.038
|
2.5261
|
2.6857
|
0.1596
|
2.0641
|
8.3968
|
23
|
h2o
|
h2o.deeplearning_first-order
|
4.372
|
2.2985
|
2.3587
|
0.0602
|
1.8349
|
7.6990
|
24
|
keras
|
fit_adadelta
|
73.920
|
4.4960
|
5.1564
|
0.6604
|
3.8748
|
16.8817
|
25
|
fit_adagrad
|
52.480
|
6.5047
|
6.7324
|
0.2277
|
5.3578
|
20.2040
|
26
|
fit_adam
|
3.620
|
2.8614
|
3.0075
|
0.1461
|
2.3994
|
8.8657
|
27
|
fit_adamax
|
6.082
|
2.5783
|
3.1920
|
0.6137
|
2.5827
|
8.6394
|
28
|
fit_nadam
|
3.166
|
5.2915
|
12.2499
|
6.9584
|
9.8464
|
27.3131
|
29
|
fit_rmsprop
|
2.552
|
6.6698
|
13.1877
|
6.5179
|
10.5806
|
32.3150
|
30
|
fit_sgd
|
13.988
|
3.4196
|
3.9121
|
0.4925
|
3.1018
|
14.8772
|
31
|
MachineShop
|
fit_none
|
0.034
|
2.2521
|
2.2681
|
0.0160
|
1.7734
|
7.5718
|
32
|
minpack.lm
|
nlsLM_none
|
0.068
|
2.2329
|
2.2329
|
0.0000
|
1.7383
|
6.9429
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.226
|
2.7246
|
5.7229
|
2.9983
|
4.8218
|
14.7964
|
34
|
monmlp.fit_Nelder-Mead
|
0.574
|
12.1868
|
12.7777
|
0.5909
|
9.8127
|
35.0537
|
35
|
neuralnet
|
neuralnet_backprop
|
0.622
|
2.9109
|
3.4962
|
0.5853
|
2.6743
|
11.9971
|
36
|
neuralnet_rprop-
|
0.308
|
2.6198
|
3.8778
|
1.2580
|
2.9818
|
13.8198
|
37
|
neuralnet_rprop+
|
0.298
|
2.9603
|
3.7997
|
0.8394
|
2.8334
|
14.1152
|
38
|
neuralnet_sag
|
7.138
|
2.2972
|
41.6253
|
39.3281
|
36.1679
|
91.5205
|
39
|
neuralnet_slr
|
0.440
|
2.8881
|
4.3263
|
1.4382
|
3.1765
|
14.4615
|
40
|
nlsr
|
nlxb_none
|
0.156
|
2.2321
|
2.6129
|
0.3808
|
2.0484
|
8.2482
|
41
|
nnet
|
nnet_none
|
0.032
|
2.2380
|
2.2557
|
0.0177
|
1.7541
|
7.3381
|
42
|
qrnn
|
qrnn.fit_none
|
0.158
|
2.7155
|
2.7208
|
0.0053
|
2.0888
|
8.1916
|
43
|
radiant.model
|
nn_none
|
0.054
|
2.3275
|
6.3010
|
3.9735
|
5.1479
|
17.4222
|
44
|
rminer
|
fit_none
|
0.084
|
2.2452
|
2.2628
|
0.0176
|
1.7485
|
7.3946
|
45
|
RSNNS
|
mlp_BackpropBatch
|
3.478
|
14.5624
|
19.9070
|
5.3446
|
16.5809
|
52.1745
|
46
|
mlp_BackpropChunk
|
0.368
|
2.9425
|
2.9820
|
0.0395
|
2.4160
|
10.5110
|
47
|
mlp_BackpropMomentum
|
0.334
|
2.8791
|
2.9074
|
0.0283
|
2.2991
|
9.6946
|
48
|
mlp_BackpropWeightDecay
|
0.372
|
2.8682
|
3.0805
|
0.2123
|
2.3657
|
8.8148
|
49
|
mlp_Quickprop
|
3.696
|
23.6323
|
24.1323
|
0.5000
|
20.1405
|
57.6774
|
50
|
mlp_Rprop
|
0.320
|
2.9859
|
10.3291
|
7.3432
|
7.2486
|
33.1339
|
51
|
mlp_SCG
|
0.558
|
2.7001
|
4.7483
|
2.0482
|
3.6042
|
16.9268
|
52
|
mlp_Std_Backpropagation
|
0.370
|
3.0594
|
3.2011
|
0.1417
|
2.5444
|
9.0758
|
53
|
snnR
|
snnR_none
|
0.038
|
11.6175
|
11.6175
|
0.0000
|
9.5749
|
25.9147
|
54
|
traineR
|
train.nnet_none
|
0.038
|
2.2431
|
2.3022
|
0.0591
|
1.7999
|
7.5122
|
55
|
validann
|
ann_BFGS
|
0.932
|
2.2606
|
2.3192
|
0.0586
|
1.8215
|
7.5409
|
56
|
ann_CG
|
41.996
|
2.3620
|
2.3962
|
0.0342
|
1.9105
|
8.2206
|
57
|
ann_L-BFGS-B
|
1.060
|
2.7065
|
3.4311
|
0.7246
|
2.6492
|
9.8990
|
58
|
ann_Nelder-Mead
|
42.914
|
9.2009
|
11.1644
|
1.9635
|
9.1979
|
26.4353
|
59
|
ann_SANN
|
0.204
|
12.6829
|
15.2562
|
2.5733
|
12.8000
|
37.0901
|
60
|
uGauss2
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.054
|
7.4794
|
8.1969
|
0.7175
|
6.0059
|
23.9130
|
1
|
train_ADAPTgdwm
|
0.068
|
4.3864
|
10.4646
|
6.0782
|
6.6054
|
38.5720
|
2
|
train_BATCHgd
|
1.552
|
9.1582
|
9.7638
|
0.6056
|
7.1783
|
28.5231
|
3
|
train_BATCHgdwm
|
1.578
|
9.2190
|
9.4697
|
0.2507
|
6.9325
|
27.9676
|
4
|
ANN2
|
neuralnetwork_adam
|
0.086
|
3.7310
|
4.0836
|
0.3526
|
3.2330
|
12.0798
|
5
|
neuralnetwork_rmsprop
|
0.100
|
3.5637
|
4.2355
|
0.6718
|
3.2536
|
15.9291
|
6
|
neuralnetwork_sgd
|
0.078
|
7.4578
|
8.3914
|
0.9336
|
6.3470
|
25.3806
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
4.952
|
8.1454
|
8.6420
|
0.4966
|
6.1349
|
30.0687
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
4.436
|
3.9723
|
8.4113
|
4.4390
|
5.6794
|
30.3739
|
9
|
automl_train_manual_trainwpso
|
8.702
|
6.1384
|
8.6501
|
2.5117
|
6.2805
|
24.3330
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.048
|
2.3781
|
3.5508
|
1.1727
|
2.9377
|
10.2283
|
11
|
CaDENCE
|
cadence.fit_optim
|
2.320
|
2.4041
|
3.1503
|
0.7462
|
2.3988
|
10.7996
|
12
|
cadence.fit_psoptim
|
5.752
|
11.1566
|
15.4210
|
4.2644
|
12.0091
|
37.0326
|
13
|
cadence.fit_Rprop
|
5.810
|
11.7899
|
17.6638
|
5.8739
|
12.7742
|
39.4495
|
14
|
caret
|
avNNet_none
|
0.086
|
2.4240
|
3.2653
|
0.8413
|
2.6499
|
9.8629
|
15
|
deepdive
|
deepnet_adam
|
0.618
|
16.8843
|
16.8843
|
0.0000
|
11.6361
|
46.6218
|
16
|
deepnet_gradientDescent
|
5.934
|
28.8118
|
28.8118
|
0.0000
|
25.1770
|
67.7823
|
17
|
deepnet_momentum
|
6.176
|
28.8104
|
28.8104
|
0.0000
|
25.1756
|
67.8020
|
18
|
deepnet_rmsProp
|
0.620
|
18.3426
|
18.3426
|
0.0000
|
12.6957
|
49.5614
|
19
|
deepnet
|
nn.train_BP
|
0.320
|
3.4032
|
6.4830
|
3.0798
|
4.8425
|
17.3581
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
20.3911
|
23.5586
|
3.1675
|
20.1337
|
51.5475
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.014
|
27.8077
|
31.2976
|
3.4899
|
25.2946
|
82.5683
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.030
|
2.7901
|
3.8821
|
1.0920
|
3.0341
|
11.9878
|
23
|
h2o
|
h2o.deeplearning_first-order
|
4.394
|
2.8574
|
3.4454
|
0.5880
|
2.7928
|
11.6466
|
24
|
keras
|
fit_adadelta
|
51.048
|
3.8515
|
3.8918
|
0.0403
|
2.8916
|
15.2710
|
25
|
fit_adagrad
|
31.862
|
5.0447
|
8.8810
|
3.8363
|
6.5481
|
25.2992
|
26
|
fit_adam
|
3.046
|
3.9732
|
6.7909
|
2.8177
|
5.1083
|
18.9390
|
27
|
fit_adamax
|
6.844
|
3.8559
|
4.2292
|
0.3733
|
3.3180
|
12.7852
|
28
|
fit_nadam
|
3.374
|
4.2819
|
7.1607
|
2.8788
|
5.6164
|
19.4085
|
29
|
fit_rmsprop
|
2.276
|
8.1016
|
10.2651
|
2.1635
|
7.5917
|
28.4689
|
30
|
fit_sgd
|
9.784
|
5.1907
|
8.8123
|
3.6216
|
6.4800
|
24.8510
|
31
|
MachineShop
|
fit_none
|
0.024
|
2.5986
|
3.1210
|
0.5224
|
2.4847
|
9.3644
|
32
|
minpack.lm
|
nlsLM_none
|
0.050
|
2.9795
|
2.9795
|
0.0000
|
2.3890
|
9.0540
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.224
|
3.0438
|
4.6769
|
1.6331
|
3.6151
|
13.7834
|
34
|
monmlp.fit_Nelder-Mead
|
0.412
|
8.2728
|
9.5898
|
1.3170
|
6.7817
|
30.1673
|
35
|
neuralnet
|
neuralnet_backprop
|
0.866
|
4.2479
|
4.3983
|
0.1504
|
3.4494
|
12.7619
|
36
|
neuralnet_rprop-
|
0.080
|
3.5904
|
4.4020
|
0.8116
|
3.4777
|
14.1560
|
37
|
neuralnet_rprop+
|
0.064
|
3.6133
|
8.6404
|
5.0271
|
6.1594
|
25.2448
|
38
|
neuralnet_sag
|
1.916
|
3.3723
|
8.6390
|
5.2667
|
6.1167
|
24.7388
|
39
|
neuralnet_slr
|
0.190
|
3.5678
|
3.7435
|
0.1757
|
2.7163
|
14.9299
|
40
|
nlsr
|
nlxb_none
|
0.118
|
2.3327
|
2.9754
|
0.6427
|
2.3838
|
9.0619
|
41
|
nnet
|
nnet_none
|
0.026
|
2.3625
|
3.0894
|
0.7269
|
2.4567
|
9.4881
|
42
|
qrnn
|
qrnn.fit_none
|
0.248
|
2.6571
|
3.7782
|
1.1211
|
2.6654
|
16.0998
|
43
|
radiant.model
|
nn_none
|
0.056
|
2.5784
|
4.5795
|
2.0011
|
3.7498
|
11.3454
|
44
|
rminer
|
fit_none
|
0.082
|
2.3604
|
2.3690
|
0.0086
|
1.8630
|
7.5477
|
45
|
RSNNS
|
mlp_BackpropBatch
|
3.266
|
12.1638
|
14.7088
|
2.5450
|
11.5655
|
30.3793
|
46
|
mlp_BackpropChunk
|
0.332
|
3.2955
|
4.7181
|
1.4226
|
3.6073
|
15.4886
|
47
|
mlp_BackpropMomentum
|
0.352
|
3.3532
|
4.8150
|
1.4618
|
3.7195
|
15.6644
|
48
|
mlp_BackpropWeightDecay
|
0.320
|
4.5703
|
6.7390
|
2.1687
|
5.5508
|
19.8255
|
49
|
mlp_Quickprop
|
3.522
|
24.5455
|
25.0662
|
0.5207
|
19.9234
|
51.7595
|
50
|
mlp_Rprop
|
0.316
|
3.5732
|
6.3892
|
2.8160
|
4.6858
|
19.1326
|
51
|
mlp_SCG
|
0.556
|
6.2398
|
6.4892
|
0.2494
|
4.8401
|
20.2438
|
52
|
mlp_Std_Backpropagation
|
0.318
|
3.4215
|
4.8320
|
1.4105
|
3.8106
|
15.2518
|
53
|
snnR
|
snnR_none
|
0.044
|
8.8419
|
9.4678
|
0.6259
|
6.9147
|
30.1105
|
54
|
traineR
|
train.nnet_none
|
0.018
|
2.3713
|
2.6020
|
0.2307
|
2.0822
|
8.0308
|
55
|
validann
|
ann_BFGS
|
0.792
|
2.3571
|
2.3654
|
0.0083
|
1.8640
|
7.5013
|
56
|
ann_CG
|
34.308
|
3.5928
|
6.3016
|
2.7088
|
4.7543
|
17.8412
|
57
|
ann_L-BFGS-B
|
0.840
|
3.0632
|
4.0776
|
1.0144
|
3.1853
|
13.2715
|
58
|
ann_Nelder-Mead
|
29.606
|
6.7221
|
7.5819
|
0.8598
|
5.9618
|
22.5839
|
59
|
ann_SANN
|
0.204
|
9.9162
|
14.2730
|
4.3568
|
11.6041
|
32.1496
|
60
|
uGauss3
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.046
|
4.7958
|
4.8043
|
0.0085
|
3.9113
|
12.5672
|
1
|
train_ADAPTgdwm
|
0.080
|
4.4658
|
5.1079
|
0.6421
|
3.6709
|
20.0320
|
2
|
train_BATCHgd
|
1.556
|
5.0863
|
5.2682
|
0.1819
|
4.1337
|
14.8772
|
3
|
train_BATCHgdwm
|
1.566
|
5.0868
|
5.2355
|
0.1487
|
4.1127
|
14.7918
|
4
|
ANN2
|
neuralnetwork_adam
|
0.082
|
3.1354
|
3.6437
|
0.5083
|
2.8900
|
10.5979
|
5
|
neuralnetwork_rmsprop
|
0.082
|
2.9727
|
3.2852
|
0.3125
|
2.5929
|
9.6420
|
6
|
neuralnetwork_sgd
|
0.080
|
4.8318
|
4.8821
|
0.0503
|
3.8567
|
14.4928
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
4.984
|
3.1214
|
3.4986
|
0.3772
|
2.7250
|
9.7689
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
4.398
|
3.5555
|
3.7519
|
0.1964
|
3.0224
|
11.8905
|
9
|
automl_train_manual_trainwpso
|
6.586
|
4.8318
|
6.6613
|
1.8295
|
4.7306
|
20.0899
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.026
|
2.8273
|
3.1966
|
0.3693
|
2.5109
|
10.0153
|
11
|
CaDENCE
|
cadence.fit_optim
|
2.334
|
2.4116
|
2.8622
|
0.4506
|
2.2233
|
7.8710
|
12
|
cadence.fit_psoptim
|
5.746
|
10.8502
|
14.8615
|
4.0113
|
11.6021
|
34.5796
|
13
|
cadence.fit_Rprop
|
5.790
|
9.1862
|
21.8896
|
12.7034
|
15.2197
|
62.4249
|
14
|
caret
|
avNNet_none
|
0.080
|
2.4976
|
3.1181
|
0.6205
|
2.3743
|
9.6517
|
15
|
deepdive
|
deepnet_adam
|
0.624
|
10.0466
|
10.0466
|
0.0000
|
7.3484
|
31.6838
|
16
|
deepnet_gradientDescent
|
5.946
|
32.2441
|
32.2441
|
0.0000
|
27.7063
|
70.1972
|
17
|
deepnet_momentum
|
6.146
|
32.2413
|
32.2413
|
0.0000
|
27.7055
|
70.2077
|
18
|
deepnet_rmsProp
|
0.622
|
20.7977
|
20.7977
|
0.0000
|
15.4531
|
49.0396
|
19
|
deepnet
|
nn.train_BP
|
0.300
|
3.5001
|
3.7035
|
0.2034
|
2.9180
|
12.1143
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
8.4445
|
19.1869
|
10.7424
|
16.8753
|
35.7678
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.010
|
31.0706
|
41.2586
|
10.1880
|
33.0051
|
112.2972
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.036
|
2.3829
|
3.1571
|
0.7742
|
2.4537
|
9.9903
|
23
|
h2o
|
h2o.deeplearning_first-order
|
4.706
|
2.9525
|
3.2931
|
0.3406
|
2.5635
|
9.9032
|
24
|
keras
|
fit_adadelta
|
19.964
|
3.9293
|
4.3921
|
0.4628
|
3.3906
|
13.1185
|
25
|
fit_adagrad
|
13.448
|
3.4760
|
4.8342
|
1.3582
|
3.7910
|
14.2616
|
26
|
fit_adam
|
2.296
|
3.2791
|
4.0840
|
0.8049
|
3.1478
|
11.9057
|
27
|
fit_adamax
|
5.008
|
2.5583
|
2.9976
|
0.4393
|
2.3357
|
9.7110
|
28
|
fit_nadam
|
2.708
|
3.2682
|
3.4600
|
0.1918
|
2.8802
|
9.8016
|
29
|
fit_rmsprop
|
1.890
|
5.4568
|
6.4457
|
0.9889
|
5.4036
|
15.4436
|
30
|
fit_sgd
|
5.274
|
4.8610
|
5.1138
|
0.2528
|
3.9761
|
15.1244
|
31
|
MachineShop
|
fit_none
|
0.034
|
2.3086
|
3.1576
|
0.8490
|
2.4940
|
9.7409
|
32
|
minpack.lm
|
nlsLM_none
|
0.040
|
3.1472
|
3.1472
|
0.0000
|
2.4837
|
9.7293
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.220
|
2.9135
|
3.5782
|
0.6647
|
2.8497
|
10.8707
|
34
|
monmlp.fit_Nelder-Mead
|
0.424
|
5.8001
|
7.3161
|
1.5160
|
5.7523
|
20.8098
|
35
|
neuralnet
|
neuralnet_backprop
|
0.302
|
3.8802
|
4.2928
|
0.4126
|
3.4361
|
11.4729
|
36
|
neuralnet_rprop-
|
0.036
|
2.6706
|
3.6253
|
0.9547
|
2.8533
|
9.9037
|
37
|
neuralnet_rprop+
|
0.066
|
2.5491
|
3.6200
|
1.0709
|
2.8585
|
10.5212
|
38
|
neuralnet_sag
|
1.016
|
2.7247
|
3.8181
|
1.0934
|
2.8290
|
11.6167
|
39
|
neuralnet_slr
|
0.102
|
2.8318
|
3.7840
|
0.9522
|
3.0024
|
10.4000
|
40
|
nlsr
|
nlxb_none
|
0.110
|
2.2991
|
2.8185
|
0.5194
|
2.2078
|
7.5077
|
41
|
nnet
|
nnet_none
|
0.020
|
2.3554
|
3.1706
|
0.8152
|
2.5057
|
9.8058
|
42
|
qrnn
|
qrnn.fit_none
|
0.208
|
2.7773
|
3.9015
|
1.1242
|
2.8959
|
13.2058
|
43
|
radiant.model
|
nn_none
|
0.046
|
2.6848
|
3.4127
|
0.7279
|
2.7026
|
10.5356
|
44
|
rminer
|
fit_none
|
0.062
|
2.3033
|
2.3232
|
0.0199
|
1.8528
|
7.0323
|
45
|
RSNNS
|
mlp_BackpropBatch
|
3.274
|
6.7104
|
9.7422
|
3.0318
|
7.6756
|
23.2263
|
46
|
mlp_BackpropChunk
|
0.320
|
2.9280
|
3.8323
|
0.9043
|
3.0061
|
11.0342
|
47
|
mlp_BackpropMomentum
|
0.320
|
2.9271
|
3.2533
|
0.3262
|
2.6246
|
8.8194
|
48
|
mlp_BackpropWeightDecay
|
0.360
|
2.8723
|
3.0215
|
0.1492
|
2.3931
|
8.5837
|
49
|
mlp_Quickprop
|
3.518
|
27.8595
|
28.8119
|
0.9524
|
22.8776
|
59.8157
|
50
|
mlp_Rprop
|
0.352
|
2.9609
|
8.8744
|
5.9135
|
6.0200
|
30.1870
|
51
|
mlp_SCG
|
0.538
|
3.3416
|
3.7196
|
0.3780
|
2.9009
|
11.2479
|
52
|
mlp_Std_Backpropagation
|
0.316
|
3.0294
|
3.2409
|
0.2115
|
2.6432
|
9.4181
|
53
|
snnR
|
snnR_none
|
0.032
|
5.2818
|
5.2818
|
0.0000
|
4.0957
|
15.6475
|
54
|
traineR
|
train.nnet_none
|
0.020
|
2.2976
|
2.8669
|
0.5693
|
2.2422
|
7.9607
|
55
|
validann
|
ann_BFGS
|
0.744
|
2.3046
|
2.8185
|
0.5139
|
2.2078
|
7.5077
|
56
|
ann_CG
|
37.490
|
2.4990
|
3.5798
|
1.0808
|
2.7264
|
11.3586
|
57
|
ann_L-BFGS-B
|
0.870
|
2.5172
|
3.5450
|
1.0278
|
2.7137
|
10.7114
|
58
|
ann_Nelder-Mead
|
30.822
|
4.9577
|
5.3229
|
0.3652
|
4.3142
|
15.0154
|
59
|
ann_SANN
|
0.210
|
6.9649
|
10.8474
|
3.8825
|
8.3651
|
26.5278
|
60
|
uNeuroOne
Package
|
Algorithm
|
Time mean
|
RMSE min
|
RMSE median
|
RMSE D51
|
MAE median
|
WAE median
|
NPFA
|
AMORE
|
train_ADAPTgd
|
0.020
|
0.2958
|
0.2965
|
0.0007
|
0.2451
|
0.6433
|
1
|
train_ADAPTgdwm
|
0.028
|
0.2854
|
0.2854
|
0.0000
|
0.2285
|
0.6436
|
2
|
train_BATCHgd
|
1.232
|
0.2931
|
0.2935
|
0.0004
|
0.2421
|
0.6309
|
3
|
train_BATCHgdwm
|
1.240
|
0.2924
|
0.2933
|
0.0009
|
0.2419
|
0.6303
|
4
|
ANN2
|
neuralnetwork_adam
|
0.008
|
0.3082
|
0.3485
|
0.0403
|
0.2776
|
0.7493
|
5
|
neuralnetwork_rmsprop
|
0.008
|
0.2904
|
0.2912
|
0.0008
|
0.2376
|
0.6015
|
6
|
neuralnetwork_sgd
|
0.010
|
0.3069
|
0.3088
|
0.0019
|
0.2535
|
0.6226
|
7
|
automl
|
automl_train_manual_trainwgrad_adam
|
1.214
|
0.2844
|
0.2895
|
0.0051
|
0.2378
|
0.6469
|
8
|
automl_train_manual_trainwgrad_RMSprop
|
1.090
|
0.2842
|
0.2888
|
0.0046
|
0.2403
|
0.6528
|
9
|
automl_train_manual_trainwpso
|
4.896
|
0.2847
|
0.2878
|
0.0031
|
0.2350
|
0.5413
|
10
|
brnn
|
brnn_Gauss-Newton
|
0.008
|
0.3523
|
0.3523
|
0.0000
|
0.2848
|
0.8271
|
11
|
CaDENCE
|
cadence.fit_optim
|
0.298
|
0.2831
|
0.2831
|
0.0000
|
0.2310
|
0.5816
|
12
|
cadence.fit_psoptim
|
4.244
|
0.5523
|
0.7577
|
0.2054
|
0.5800
|
1.9676
|
13
|
cadence.fit_Rprop
|
2.786
|
0.3054
|
0.3248
|
0.0194
|
0.2626
|
0.7872
|
14
|
caret
|
avNNet_none
|
0.010
|
0.2904
|
0.2946
|
0.0042
|
0.2437
|
0.6434
|
15
|
deepdive
|
deepnet_adam
|
0.562
|
0.2946
|
0.2946
|
0.0000
|
0.2459
|
0.5582
|
16
|
deepnet_gradientDescent
|
5.160
|
0.3666
|
0.3666
|
0.0000
|
0.3105
|
0.6748
|
17
|
deepnet_momentum
|
5.348
|
0.3544
|
0.3544
|
0.0000
|
0.3001
|
0.6152
|
18
|
deepnet_rmsProp
|
0.552
|
0.3161
|
0.3161
|
0.0000
|
0.2695
|
0.5981
|
19
|
deepnet
|
nn.train_BP
|
0.084
|
0.2830
|
0.2830
|
0.0000
|
0.2314
|
0.5653
|
20
|
elmNNRcpp
|
elm_train_extremeML
|
0.000
|
0.8650
|
0.9526
|
0.0876
|
0.7905
|
2.2943
|
21
|
ELMR
|
OSelm_train.formula_extremeML
|
0.000
|
0.9735
|
1.0466
|
0.0731
|
0.8640
|
2.4817
|
22
|
EnsembleBase
|
Regression.Batch.Fit_none
|
0.012
|
0.2826
|
0.2831
|
0.0005
|
0.2326
|
0.5543
|
23
|
h2o
|
h2o.deeplearning_first-order
|
3.344
|
0.2831
|
0.2832
|
0.0001
|
0.2331
|
0.5539
|
24
|
keras
|
fit_adadelta
|
19.378
|
0.2871
|
0.2879
|
0.0008
|
0.2377
|
0.5887
|
25
|
fit_adagrad
|
13.868
|
0.2893
|
0.2936
|
0.0043
|
0.2429
|
0.5637
|
26
|
fit_adam
|
1.320
|
0.2869
|
0.2875
|
0.0006
|
0.2340
|
0.5886
|
27
|
fit_adamax
|
2.502
|
0.2841
|
0.2864
|
0.0023
|
0.2366
|
0.5789
|
28
|
fit_nadam
|
1.518
|
0.2855
|
0.2896
|
0.0041
|
0.2437
|
0.6055
|
29
|
fit_rmsprop
|
1.108
|
0.3042
|
0.3629
|
0.0587
|
0.3049
|
0.7486
|
30
|
fit_sgd
|
3.468
|
0.2901
|
0.2922
|
0.0021
|
0.2410
|
0.5769
|
31
|
MachineShop
|
fit_none
|
0.010
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
32
|
minpack.lm
|
nlsLM_none
|
0.004
|
1.2720
|
1.2720
|
0.0000
|
1.1104
|
2.5150
|
33
|
monmlp
|
monmlp.fit_BFGS
|
0.194
|
0.2831
|
0.2834
|
0.0003
|
0.2312
|
0.5810
|
34
|
monmlp.fit_Nelder-Mead
|
0.224
|
0.3020
|
0.3266
|
0.0246
|
0.2601
|
0.6451
|
35
|
neuralnet
|
neuralnet_backprop
|
0.152
|
0.2898
|
0.2926
|
0.0028
|
0.2423
|
0.5889
|
36
|
neuralnet_rprop-
|
0.010
|
0.2864
|
0.2935
|
0.0071
|
0.2480
|
0.6059
|
37
|
neuralnet_rprop+
|
0.000
|
0.2848
|
0.3165
|
0.0317
|
0.2586
|
0.6196
|
38
|
neuralnet_sag
|
0.056
|
0.2893
|
0.3212
|
0.0319
|
0.2638
|
0.6316
|
39
|
neuralnet_slr
|
0.052
|
0.2923
|
0.3203
|
0.0280
|
0.2607
|
0.8073
|
40
|
nlsr
|
nlxb_none
|
0.008
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
41
|
nnet
|
nnet_none
|
0.000
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
42
|
qrnn
|
qrnn.fit_none
|
0.094
|
0.2939
|
0.2939
|
0.0000
|
0.2258
|
0.7231
|
43
|
radiant.model
|
nn_none
|
0.010
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5677
|
44
|
rminer
|
fit_none
|
0.004
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
45
|
RSNNS
|
mlp_BackpropBatch
|
0.788
|
0.6867
|
0.6888
|
0.0021
|
0.5629
|
1.6534
|
46
|
mlp_BackpropChunk
|
0.074
|
0.2912
|
0.6365
|
0.3453
|
0.5156
|
1.6363
|
47
|
mlp_BackpropMomentum
|
0.074
|
0.2968
|
0.3315
|
0.0347
|
0.2742
|
0.7631
|
48
|
mlp_BackpropWeightDecay
|
0.082
|
0.3096
|
0.6423
|
0.3327
|
0.5179
|
1.6618
|
49
|
mlp_Quickprop
|
0.764
|
0.5304
|
0.5304
|
0.0000
|
0.4235
|
1.2829
|
50
|
mlp_Rprop
|
0.076
|
0.2830
|
0.3141
|
0.0311
|
0.2531
|
0.7252
|
51
|
mlp_SCG
|
0.104
|
0.2855
|
0.6216
|
0.3361
|
0.5100
|
1.4782
|
52
|
mlp_Std_Backpropagation
|
0.082
|
0.2834
|
0.3135
|
0.0301
|
0.2457
|
0.7675
|
53
|
snnR
|
snnR_none
|
0.004
|
0.6793
|
0.6793
|
0.0000
|
0.5564
|
1.6288
|
54
|
traineR
|
train.nnet_none
|
0.000
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
55
|
validann
|
ann_BFGS
|
0.104
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
56
|
ann_CG
|
23.762
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
57
|
ann_L-BFGS-B
|
0.222
|
0.2830
|
0.2830
|
0.0000
|
0.2313
|
0.5675
|
58
|
ann_Nelder-Mead
|
9.006
|
0.3256
|
0.3341
|
0.0085
|
0.2793
|
0.8397
|
59
|
ann_SANN
|
0.168
|
0.3084
|
0.3344
|
0.0260
|
0.2773
|
0.6937
|
60
|
Summary statistics for top-10 packages on bWoodN1 Dataset
The table provides the summary statistics of top-10 NN packages over 20 runs on the large dataset bWoodN1
which contains 20,000 rows with 6 inputs valued in [0,1] for a (single) numeric output.
Package
|
Time (mean)
|
RMSE (min)
|
RMSE (median)
|
RMSE (D51)
|
MAE (median)
|
WAE (median)
|
rminer
|
10.2890
|
3.3662
|
3.56120
|
0.19500
|
2.86775
|
14.83000
|
traineR
|
3.2575
|
3.5488
|
4.57530
|
1.02650
|
3.72975
|
15.38630
|
CaDENCE
|
228.5210
|
3.3667
|
4.60750
|
1.24080
|
3.78875
|
16.57340
|
validann
|
145.0263
|
3.3800
|
4.62390
|
1.24390
|
3.75110
|
16.36570
|
h2o
|
127.9933
|
4.5704
|
4.64945
|
0.07905
|
3.76315
|
17.46895
|
monmlp
|
8.6500
|
4.5442
|
4.70060
|
0.15640
|
3.80100
|
15.31885
|
nlsr
|
73.3358
|
3.5512
|
4.70250
|
1.15130
|
3.80060
|
16.72755
|
nnet
|
3.4488
|
3.5499
|
4.70570
|
1.15580
|
3.79840
|
16.51300
|
MachineShop
|
3.6589
|
3.5518
|
4.77470
|
1.22290
|
3.84850
|
15.30465
|
radiant.model
|
0.0036
|
10.9572
|
10.95720
|
0.00000
|
8.76850
|
42.61880
|