Particle identification using machine learning at the HADES experiment

The recent outburst in the popularity of the computational technology associated with a broad topic of data science results in a high and still growing accessibility of tools allowing one to make use of sophisticated machine learning algorithms without an expert-level knowledge in the field. Due to that, the usage of such mechanisms became widespread in many branches of business and science, where it helps to analyse large and/or complicated data sets in a relatively quick and efficient way, while obtaining highly accurate results. The presentation briefly describes the preparation, evaluation and usage of a solution based on these mechanisms and designed for the problem of particle identification in the HADES experiment at the Facility for Antiproton and Ion Research in Europe.
The research was conducted with the usage of data coming from 1.23 AGeV Au+Au collisions in the HADES detector simulated using Monte Carlo methods of the GEANT toolkit. The next step should focus on applying the prepared solution to experimental data, but the development is still ongoing.