## Techniques for Analyzing Large Numbers of Light Curves

The critical issues are speed and stability. Speed is obviously necessary to analyze large numbers of data. Stability is required to automate the procedure. Automation is required if the user is to analyze large sets of data produced by surveys. Approaches to automating light curve solutions have taken various forms to date.

1. Matching approach [Wyithe & Wilson 2002a,b and Wilson & Wyithe (2003)]: Match one or several light curves to a large test set of pre-computed light curves. This set is used in the first step to analyze observed light curves.

2. Rule-based approaches: These follow an a priori rule-based procedure to extract relevant light curve information to produce a good initial parameter set for further light curve fitting. They are helpful to non-experts and can be combined with the matching approach.

3. Simplified physical models: Devor (2005) developed an automated pipeline for a simple spherical star model without tidal or reflection physics, whose starting

6 Methods on how eclipsing binary stars are detected in surveys and distinguished from other variable stars are found, for instance, in Eyer & Blake (2005) and references therein.

values are guessed and then refined with a downhill simplex method followed by simulated annealing. This approach identifies detached binaries which can be described sufficiently accurate by spherical models. Tamuz et al. (2006) employed the EBOP ellipsoidal model (Popper & Etzel 1981). Using this engine, they arrived at initial solutions after a combination of grid search, gradient descent, and geometrical light curve analysis. 4. Artificial neural network (Prsa et al. (2008)): The neural network approach can be understood as a formal mathematical approximation technique in which amplifiers between input and output signals are adjusted.

The first two approaches have in common that they can exploit the correct binary star physics that gives rise to the light curves. The neural network approach as a formal approximation techniques may lack this facet without extra information. If spherical or ellipsoidal stars are used to approximate semi-detached or over-contact binaries, it is not clear if the initial information obtained from such simplifications will be useful. All of these treatments have the advantage that they allow non-experts to produce reasonable initial parameter sets for further detailed study.

A prerequisite for the first two approaches is to know the ephemerides. In principle, the period may be extracted by a power series analysis. Surveys may not have a long-enough duration to derive period changes. However, period changes might be an issue for non-expert users when analyzing EB light curves obtained at different epochs.

## Telescopes Mastery

Through this ebook, you are going to learn what you will need to know all about the telescopes that can provide a fun and rewarding hobby for you and your family!

## Post a comment