For large-scale air pollution monitoring data, this study proposed an effective interval grey number approximation aggregation method for time series and designed hierarchical clustering algorithm to help develop air pollution control strategies.
This study first introduces the Kernel Density Estimation (KDE) method to estimate the value distribution of the raw time series. Then, Multi-Swap k-Means++ clustering is applied to achieve optimal segmentation of the series. The segmented time series is represented using interval grey number sequences. Finally, a hierarchical clustering algorithm is developed based on the GAX (the interval Grey number Aggregation approximation) representation to cluster the time series effectively.
The method proposed in this paper implements effective compression and simplified representation of time series data, which improves the accuracy and efficiency of large-scale time series clustering analyses and provides solid technical support for the monitoring and management of air pollutants.
Through in-depth analysis of air pollution monitoring data, the methodology of this paper helps to discover the existence mode and potential structure of pollutants, scientifically formulate effective pollution control strategies, which helps to improve air quality and protect public health.
