A Stock Selection Model of Image Classification Method Based on Convolutional Neural Network

Algorithm 1

Factor selection.

	Input: Data set: ; % is the primitive factor number
	Procedure:
(1)	For:
(2)	For:
(3)	Calculate the correlation of factor and
(4)	End For
(5)	End For
(6)	Delete 50% of the factors with lower correlation, the remaining number of factor is n
(7)	For:
(8)	Calculate the correlation of factor and , is the return in the next month
(9)	End For
(10)	Delete 50% of the factors with higher correlation
	Output: Multifactor stock selection data set