Table 1 Table showing the fixed parameters during evaluation, and the parameters, which are optimized by using grid search

MLR

-

-

-

SVR

RBF kernel w. γ=“scale”

C=[1,...,80],

C=30,

ε =[0.008,...,0.032]

ε=0.016

K-NN

weights="uniform"

p=[1,2,3],

p=1,

K-D tree

k=[1,...,80]

k=40

Decision Tree

min_samples_leaf=5

N_leave=[100,...,30000]

N_leave=10000

Random Forest

N_leave=10000,

N_trees=[5,...,140]

N_trees=20

AdaBoost

N_leave=10000,

N_trees=[5,...,140],

N_trees=40,

loss=“linear”

η=[0.5,...,2]

η=2

GBR

N_leave=31,η=0.1,

N_trees=[60,...,900]

N_trees=480

loss=“least squares”

FFNN

activation=“ReLU”

L_FC=[1,2,3,4],

L_FC=3,

N_FC=[8,...,100]

N_FC=48

MergeFFNN

activation=“ReLU”

N_top=[5,...,50],

N_top=35,

N_FC,a.m.=N_top+N_bot

N_bot=[10,...,40]

N_bot=30

MergeLSTM

N_FC,a.m.=N_LSTM+N_FC

N_LSTM=[5,...,30],

N_LSTM=5,

N_FC=[15,...,40]

N_FC=25

LSTM

-

N_LSTM=[5,...,35],

N_LSTM=15,

N_FC=[0,...,40]

N_FC=40

MergeCNN

N_FC,a.m.=N_filter+N_FC

N_filter=[5,...,25],

N_filter=15,

filter_size,conv_Stride

N_FC=[10,...,40]

N_FC=40

CNN

filter_size,conv_Stride

N_filter=[5,...,40],

N_filter=25,

N_FC=[0,...,50]

N_FC=35