Aeronautical γ Spectrum Noise Reduction Method Based on LS-SVM Segmentation Regression

In order to reduce noise of the aaeronautical γ spectrum caused by factors such as less counting, short measuring time and poor measuring environment, this article adopts a segmented noise reduction method combined with the machine learning method, LS-SVM(Least Squares Support Vector Machines) and weighted stacking method according to the energy window distribution of the spectrum. The experimental results show that the segmented noise reduction method based on LS-SVM can significantly reduce the statistical fluctuation. In addition, the method presents good adaptability and generalization ability.