Application of Multivariate Analysis Methods to Search for Variable Stars in the Pi of the Sky experiment

Pi of the Sky is a system of wide field-of-view robotic telescopes, which
search for short timescale astrophysical phenomena, especially for prompt
optical emission from GRB. The system was designed for autonomous
operation, monitoring a large fraction of the sky with 12m -13m range and
time resolution of the order of 1 - 100 seconds.

Dedicated photometric algorithm based on ASAS was implemented in the LUIZA
framework and optimized on a sample of test sky images. It was then
applied on a "training" image sample, which was obtained by modifying the
test sample of images by applying different patterns of variability to
selected constant stars. Different statistical estimators were then
considered to find the most efficient algorithm, based on Multivariate
Analysis (MVA) method, for variable star identification. Analysis of
the test results indicated that most efficient candidate star selection
can be based on the so called Boosted Decision Tree (BDT) approach.
The algorithm is then used to search variable star candidates in the
actutal data.