An Empirical Study of Machine Learning Algorithms for Stock Daily Trading Strategy

Table 4

Trading performance of different trading strategies in the SPICS. Best performance of all trading strategies is in boldface.

	Index	BAH	MLP	DBN	SAE	RNN	LSTM	GRU	CART	NB	RF	LR	SVM	XGB
AR	—	—	0.5205	0.5189	0.5201	0.5025	0.5013	0.4986	0.6309	0.5476	0.6431	0.6491	0.6235	0.6600
PR	—	—	0.7861	0.7764	0.7781	0.5427	0.5121	0.4911	0.6514	0.5270	0.6595	0.6474	0.6733	0.6738
RR	—	—	0.5274	0.5263	0.5273	0.5245	0.5253	0.5239	0.6472	0.5762	0.6599	0.6722	0.6325	0.6767
F1	—	—	0.6258	0.6217	0.6229	0.5332	0.5183	0.5065	0.6491	0.5480	0.6595	0.6591	0.6517	0.6751
AUC	—	—	0.5003	0.5001	0.5002	0.4997	0.5005	0.4992	0.6295	0.5489	0.6418	0.6491	0.6199	0.6590
WR	0.5450	0.5235	0.5676	0.5680	0.5683	0.5843	0.5825	0.5844	0.5266	0.5930	0.5912	0.5859	0.5831	0.5891
ARR	0.1227	0.1603	0.3333	0.3298	0.3327	0.2945	0.2921	0.2935	0.3319	0.2976	0.3134	0.2944	0.3068	0.3042
ASR	0.8375	0.6553	1.5472	1.5415	1.5506	1.5768	1.5575	1.5832	1.3931	1.6241	1.6768	1.5822	1.6022	1.6302
MDD	0.1939	0.4233	0.3584	0.3585	0.3547	0.3403	0.3489	0.3381	0.3413	0.3428	0.3284	0.3447	0.3429	0.3338