Missing data simulation method based on degradation trend analysis and GRNN

• Main Authors: FAN YE, DE XINLIN, YANG LU, ZHOU JIE, LIU YUFENG, YANG XUEYIN, WANG QIANG
• Format: Patent
• Language: Chinese; English
• Published: 17.01.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention provides a missing data simulation method based on degradation trend analysis and a GRNN, and the method comprises the steps: firstly carrying out the analysis and modeling of an observed degradation data trend, then building a GRNN neural network model of a degradation data residual error sequence, and estimating the residual error sequence of missing data; and finally, integrating the data degradation trend and the estimation result of the residual error sequence to obtain a simulation value of missing data. According to the missing data simulation method based on the degradation trend analysis and the GRNN, the missing data is simulated by comprehensively collecting trend analysis and the GRNN, discreteness and volatility of original data can be well restored, meanwhile, the missing data simulation method based on the degradation trend analysis and the GRNN has higher advantages in the approximation ability and the learning speed, and the missing data simulation method based on the degradatio