Factor analysis of COVID-19 in China based on modal decomposition and independent component analysis
Zhiying Hou, Jimin Ye · 2023
Since December 2019, the novel coronavirus disease 2019 (COVID-19) has erupted around the world, seriously affecting people’s lives, hindering economic development, and threatening public health security. As a result, it is critical to investigate the COVID-19 impacting elements to take the required steps to deal with the current crisis and minimize potential hazards in the future. Based on data from daily confirmed COVID-19 cases in China, this data series is first decomposed into multiple intrinsic modal functions (IMFs) using modal decomposition, and then the IMFs are restructured according to the relative hamming distance (RHD). To investigate the influence of factors internal to COVID-19, statistically independent sub-series (ICs) are isolated from the restructured modalities using independent component analysis (ICA). In addition, it is known that temperature, the consumer expectation index (CEI), the migration scale index (MSI), and the month-on-month growth of the consumer price index (CPI) have an impact on COVID-19. Finally, the analysis shows that the trend between ICs and these factors is highly similar, thus identifying several relatively independent main influencing factors of COVID-19 from the data itself.