Search for low mass dark matter in DarkSide-50: the bayesian network approach

We present a novel approach for the search of dark matter in the DarkSide-50experiment, relying on Bayesian Networks. This method incorporates the detectorresponse model into the likelihood function, explicitly maintaining theconnection with the quantity of interest. No assumptions about the linearity ofthe problem or the shape of the probability distribution functions arerequired, and there is no need to morph signal and background spectra as afunction of nuisance parameters. By expressing the problem in terms of BayesianNetworks, we have developed an inference algorithm based on a Markov ChainMonte Carlo to calculate the posterior probability. A clever description of thedetector response model in terms of parametric matrices allows us to study theimpact of systematic variations of any parameter on the final results. Ourapproach not only provides the desired information on the parameter ofinterest, but also potential constraints on the response model. Our results areconsistent with recent published analyses and further refine the parameters ofthe detector response model.