Gas Concentration Monitoring Prewarning Based on Adaptive Prediction and Feature Extraction

Table 2

Comparison with other prediction methods.

Relevant research	Prediction model	Implementation process	Prediction effectiveness
[1]	Chaotic time series model	Phase space reconstruction	0.69
[2]	Support vector machine regression	Fix the time length to reconstruct the sequence	0.77
[3]	Deep learning	Dynamically adjust the time length to reconstruct the sequence	0.86
[4]	Recurrent neural network	Fix the time length to reconstruct the sequence	0.78
[5]	Grey self-memorization	Grey prediction model	0.71
[6]	Extreme learning machine	Dynamically adjust the time length to reconstruct the sequence	0.83
	This research method	Multiparameter adaptive adjustment driven by real-time data	0.92