Identifying Galaxy Cluster Mergers with Deep Neural Networks using Idealized Compton-y and X-ray maps

We present a novel approach to identify galaxy clusters that are undergoing a merger using a deep learning approach. This paper uses massive galaxy clusters spanning $0 \leq z \leq 2$ from \textsc{The Three Hundred} project, a suite of hydrodynamic re-simulations of 324 large galaxy clusters. Mock, idealised Compton-{\it y} and X-ray maps were constructed for the sample, capturing them out to a radius of $2R_{200}$. The idealised nature of these maps mean they do not consider observational effects such as foreground or background astrophysical objects, any spatial resolution limits or restriction on X-ray energy bands. Half of the maps belong to a merging population as defined by a mass increase $\Delta${\it M/M} $\geq$ 0.75, and the other half serve as a control, relaxed population. We employ a convolutional neural network architecture and train the model to classify clusters into one of the groups. A best-performing model was able to correctly distinguish between the two populations with a balanced accuracy (BA) and recall of 0.77, ROC-AUC of 0.85, PR-AUC of 0.55 and $F_{1}$ score of 0.53. Using a multichannel model relative to a single channel model, we obtain a 3\% improvement in BA score, and a 6\% improvement in $F_{1}$ score. We use a saliency interpretation approach to discern the regions most important to each classification decision. By analysing radially binned saliency values we find a preference to utilise regions out to larger distances for mergers with respect to non-mergers, greater than $\sim1.2 R_{200}$ and $\sim0.7 R_{200}$ for SZ and X-ray respectively.