A Hybrid Model Using PCA and BP Neural Network for Time Series Prediction in Chinese Stock Market with TOPSIS Analysis

Table 1

Comparison of the proposed approach to existing studies.

Literature	Method			Evaluation indexes
	Algorithm	Exponential smoothing	Hybrid model	MSE	RMSE	MAPE	Accuracy	Accuracy5	TOPSIS
[14]	ANN	No	No	No	No	Yes	No	No	No
[15]	BP.	No	No	No	No	No	No	No	No
[16]	ANN	No	ARIMA-ANN	Yes	Yes	No	No	No	No
[17]	BP	No	No	Yes	Yes	No	No	No	No
[18]	MLP/LSTM	No	No	Yes	Yes	No	No	No	No
[19]	LSTM	No	No	No	Yes	No	No	No	No
[20]	LSTM	No	No	Yes	Yes	No	No	No	No
[21]	LSTM	No	No	Yes	Yes	Yes	No	No	No
[22]	CNN	No	No	No	No	No	No	No	No
[23]	CNN	No	No	No	No	No	No	No	No
[24]	CNN/LSTM	No	CNN-LSTM	No	No	No	No	No	No
[25]	CEEMD	No	CEEMD-Entropy-GRU-HA	No	Yes	No	No	No	No
Proposed approach	Yes	Yes	PCA–BP	Yes	Yes	Yes	Yes	Yes	Yes

RMSE, root mean square error.