A Multiple Kernel Learning Approach for Air Quality Prediction

Table 5

Main parameters and their tuning range of the used algorithms.

Algorithm	Parameter	Algorithm	Parameter
ARIMA	p: [0,3], d: [0,10], q: [0,3]
RF	n_estimators: [100, 1000; 50]; max_depth: [10, 20; 1]; max_features: [10, 30; 1]; min_samples_split: [2,100; 1]; min_samples_leaf: [1,100; 1]	SVC_linear	C: [100, 5000; 100]
		SVC_rbf	C: [100, 5000; 100]; gamma: [0.0, 1.0; 0.01]
MLP	hidden_layer_sizes: {(50, 50), (100, 100), (10, 20, 10), (20, 40, 20)}; activation: {‘identity’, ‘logistic’,‘tanh’, ‘relu’}; solver: {‘lbfgs’, ‘sgd’, ‘adam’}	SVC_sig	C: [100, 5000; 100]; gamma: [0.0, 1.0; 0.01]; coef0: [0, 1000; 50]
		SVC_poly	C: [100, 5000; 100]; degree: {2,3}; gamma: [0.0, 1.0; 0.01]; coef0: [0, 1000; 50]

p: AR specification; d: integration order; q: MA specification; C: regularization coefficient in SVC; n_estimators: number of trees in the forest; max_depth: maximum depth of the tree; max_features: maximum number of features when looking for the best split; min_samples_split: the minimum number of samples required to split an internal node; min_samples_leaf: the minimum number of samples required to be at a leaf node; solver: algorithm used in the optimization problem; hidden_layer_sizes: hidden layer size; alpha: regularization term parameter in MLP; activation: activation function for the hidden layer; gamma: kernel coefficient for ‘rbf’, ‘poly’, and ‘sigmoid’; degree: degree of the polynomial kernel function; coef0: independent term in kernel functions for ‘poly’ and ‘sigmoid’; [a, b; c] means within range [a, b], increase c every iteration; {} means set of values.