Socially Responsible Investment Portfolio Construction with a Double-Screening Mechanism considering Machine Learning Prediction

Table 2

Parameter settings of different models.

Prediction models	Parameter	Value
SVR	Kernel function
	C
	DNN	Layer structure (number of neurons)
Activation function		Relu
Dropout		0.2
Learning rate		0.01
The number of epochs		200
LSTM	Layer structure (number of neurons)
	Dropout	0.1
	Learning rate	0.01
	Number of epochs	200
ELM	Activation function
	Neuron number of hidden layer	100
GA	Population size
	Maximum generation
	Crossover probability
	Mutation probability