Abstract: In recent years, data preprocessing and utilization of natural environmental γ radiation dose rate have become one of the hot spots in the field of environmental quality monitoring. This paper proposes data preprocessing procedures through statistical data analysis, data cleaning, and time sequences denoising. And this paper also investigates the impact of data preprocessing with evaluation to provide high-quality data for outliers detection techniques based on long short-term memory (LSTM). After data preprocessing, data quality is improved, and the outliers detection is significantly improved in terms of accuracy, precision, recall, and F1-score ect. Data preprocessing lays a good foundation for further data mining and research of outliers.

Key words: data preprocessing, natural environmental γ radiation dose rate, time sequences, outliers detection, LSTM