[Data Mining for CVs Spectra in LAMOST-DR1].

Area of Science:

• Astronomy and Astrophysics
• Data Science

Background:

• The Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST) provides the world's largest collection of stellar spectra.
• Identifying rare celestial objects like cataclysmic variable stars requires advanced data processing techniques.

Purpose of the Study:

• To develop and apply a data mining method for detecting cataclysmic variable star spectra within the massive LAMOST-DR1 dataset.
• To validate the feasibility of the proposed approach using real astronomical data.

Main Methods:

• Utilized Laplacian Eigenmap for constructing a feature space to differentiate celestial spectra.
• Employed particle swarm optimization to tune artificial neural network parameters.
• Processed the entire LAMOST-DR1 data to search for target spectra.

Main Results:

• Successfully identified 7 cataclysmic variable star spectra from the LAMOST-DR1 data.
• The findings include 2 dwarf novae, 2 nova-like variables, and one highly polarized AM Her type variable.
• The newly discovered spectra significantly contribute to the existing library of cataclysmic variable spectra.

Conclusions:

• The developed data mining method is effective and feasible for analyzing LAMOST data.
• This approach demonstrates potential for discovering other rare celestial objects in large astronomical surveys.
• The study represents the first successful search for cataclysmic variable stars using LAMOST data.