Statistical Data Analysis Methods

The framework for the search for new very-high-energy emissions through the log-likelihood ratio test statistics method on the measured datasets is described. A quick overview of standard Test-Statistics (TS) based methods is given followed by a discussion on how these methods are applied to derive limits on the free parameters of the tested model for the searched emission. Nuisance parameters are introduced to account for the systematic uncertainties in these methods and their impact on the final limits are discussed. The performances expected to reconstruct injected fake signals in a measured background dataset are discussed together with the method to build TS profiles for limit computation with a combination of independent datasets.