Spectral methods for imputation of missing air quality data

摘要: Air quality is well recognized as a contributing factor for various physical phenomena and public health risk factor. Consequently, there need an accurate way to measure the level of exposure pollutants. Longitudinal continuous monitoring however, often incomplete due measurement errors, hardware problems or insufficient sampling frequency. In this paper we introduce discrete theorem task imputing missing data in longitudinal air-quality time series. Within context theorem, two spectral schemes filling values are presented—a Discrete Cosine Transform (DCT) Clustering Single Variable Decomposition (K-SVD) based methods. The evaluation suggested methods terms accuracy robustness showed that comparable state art when at random do have upper hand big chunks. was evaluated using complete very long air pollutants Previous studies used shorter series, altering results. imputation method by examining its performance with increasing portions data. Spectral great option imputation, which should be considered especially patterns unknown.