First application of machine learning algorithms to the position reconstruction in Resistive Silicon Detectors

Abstract RSDs (Resistive AC-Coupled Silicon Detectors) are n-in-p silicon sensors based on the LGAD (Low-Gain Avalanche Diode) technology, featuring a continuous gain layer over whole sensor area. The truly innovative feature of these is that signal induced by an ionising particle seen several pixels, allowing use reconstruction techniques combine information from many read-out channels. In this contribution, first application machine learning technique to RSD devices presented. spatial resolution compared obtained with standard methods analytical descriptions sharing mechanism. A Multi-Output regressor algorithm, trained combination simulated and real data, leads less than 2 ?m for 100 pixel. prospects future improvements also discussed.