Abstract
The paper describes a new fast and efficient algorithm to classify waveforms from signals that arise in gravitational wave detectors like LIGO. Such waveform classification is useful and important for detector characterization as well as for understanding glitches seen in the analysis pipelines that detect signals from astrophysical burst and compact object inspiral sources. Classification of glitches based on discrete parameters has been reported earlier by the author. In the current study, a new feature-mining approach has been developed that uses the additional information based on shape of the glitch waveforms. This has been a challenging problem because of the unique structures present in the real gravitational wave data. The paper presents results from simulations as well as real data. Studies show that the proposed method is fast and efficient in classifying glitches in noise.