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

Table 1

Results of decision tree under different parameters.

Training size	Leaf nodes	Depth	Benchmark	Annual return	Sharp ratio	Max drawdown
24	50	20	29864	15956	−0.0073	0.1946
36	50	20	45285	53892	0.02464	0.2008
48	50	20	35711	2871	0.0050	0.2199
24	50	25	29864	7042	−0.0027	0.2186
36	50	25	45285	57700	0.0236	0.2072
48	50	25	35711	33672	−0.0072	0.2689
24	50	30	29864	9193	−0.0083	0.1953
36	50	30	45285	44739	0.0168	0.2077
48	50	30	35711	5885	0.0078	0.2010
24	30	25	29864	43062	0.0059	0.2087
36	30	25	45285	35755	0.0145	0.2244
48	30	25	35711	5259	0.0051	0.2631
24	50	25	29864	10775	−0.0020	0.2065
36	50	25	45285	49842	0.0164	0.1878
48	50	25	35711	177	0.0092	0.2265
24	60	25	29864	2796	−0.0136	0.2071
36	60	25	45285	35754	0.0101	0.2066
48	60	25	35711	10825	0.0134	0.2355