Algorithm

Optimized Parameters

Total algorithms tested

Best Parameters combination

SVR

C, ξ, ϒ

72

C = 5, ξ = 0.00001, ϒ = scale

Regression

Tree

α, min samples split, min samples leaf, min weight fraction, splitter

900

α = 1,

min samples leaf = 1,

min samples split = 3,

min weight fraction = 0,

splitter = random

Random Forest

min samples leaf, min samples split, number of trees

1250

min samples leaf = 1,

min samples split = 2,

number of trees = 2375

ANN

batch size, epochs, no. of neurons, no. of hidden layers, weights initialization, activation, optimization algorithm

40,000

batch size = 8,

epochs = 150,

no. of neurons hidden layer = 141,

no. of hidden layers = 2,

weights initialization = Xavier Uniform,

activation = RELU,

optimization = Adam