Data Calibration Based on Multisensor Using Classification Analysis: A Random Forests Approach

Table 1

Properties Description of the Non-equilibrium Datasets.

Data Set	Data	Sample size	Outlier size	Normal sample	Non-equilibrium rate
Jibei Station; Data Set	Induction loop data 1	4824	182	4642	3.921%
	Magnetic data 1	6399	443	5956	7.452%
	Monitoring data 1	10870	659	9619	6.851%
Jiaozhou Station; Data Set	Induction loop data 2	5475	201	5217	3.671%
	Magnetic data 2	6001	342	5633	5.699%
	Monitoring data 2	12801	779	11939	6.085%
Gaotang Station; Data Set	Induction loop data 3	5013	198	4762	3.949%
	Magnetic data 3	6512	531	5972	8.154%
	Monitoring data 3	11194	625	10478	5.583%