Applying the dual-tree complex DWT and double density DWT for mass spectrometry feature extraction and classification

摘要

In this paper, we propose to use dual-tree complex and double density discrete wavelet transform for extracting mass spectrometry features. Two corresponding procedures are suggested for mass spectrometry classification. Several experiments are deployed on two types of MALDI-TOF mass spectra, including stable spectra and noisy spectra. The classification results show that our proposed procedures not only obtain better performance than previous methods but are robust-to-noise.