We present a machine learning (ML) pipeline to identify star clusters in the multicolor images of nearby galaxies, from observations obtained with the Hubble Space Telescope as part of the Treasury Project LEGUS (Legacy ExtraGalactic Ultraviolet Survey). StarcNet (STAR Cluster classification NETwork) is a multiscale convolutional neural network (CNN) that achieves an accuracy of 68.6% (four classes)/86.0% (two classes: cluster/noncluster) for star cluster classification in the images of the LEGUS galaxies, nearly matching human expert performance. We test the performance of StarcNet by applying a pre-trained CNN model to galaxies not included in the training set, finding accuracies similar to the reference one. We test the effect of StarcNet predictions on the inferred cluster properties by comparing multicolor luminosity functions and mass-age plots from catalogs produced by StarcNet and by human labeling; distributions in luminosity, color, and physical characteristics of star clusters are similar for the human and ML classified samples. There are two advantages to the ML approach: (1) reproducibility of the classifications: the ML algorithm's biases are fixed and can be measured for subsequent analysis; and (2) speed of classification: the algorithm requires minutes for tasks that humans require weeks to months to perform. By achieving comparable accuracy to human classifiers, StarcNet will enable extending classifications to a larger number of candidate samples than currently available, thus increasing significantly the statistics for cluster studies.

The relative average minimum projected separations of star clusters in the Legacy ExtraGalactic UV Survey (LEGUS) and in tidal dwarfs around the interacting galaxy NGC 5291 are determined as a function of cluster mass to look for cluster-cluster mass segregation. Class 2 and 3 LEGUS clusters, which have a more irregular internal structure than the compact and symmetric class 1 clusters, are found to be mass-segregated in low-mass galaxies, which means that the more massive clusters are systematically bunched together compared to the lower-mass clusters. This mass segregation is not present in high-mass galaxies or class 1 clusters. We consider possible causes for this segregation, including differences in cluster formation and scattering in the shallow gravitational potentials of low-mass galaxies.

We present results from a joint ALMA/HST study of the nearby spiral galaxy NGC.628. We combine the Hubble Space Telescope (HST) Legacy ExtraGalactic UV Survey (LEGUS) database of over 1000 stellar clusters in NGC.628 with ALMA Cycle 4 mm/submillimeter observations of the cold dust continuum that span similar to 15.kpc(2) including the nuclear region and western portions of the galaxy's disk. The resolution-1 ''.1 or approximately 50 pc at the distance of NGC.628-allows us to constrain the spatial variations in the slope of the millimeter dust continuum as a function of the ages and masses of the nearby stellar clusters. Our results indicate an excess of dust emission in the millimeter, assuming a typical cold dust model for a normal star-forming galaxy, but little correlation of the dust continuum slope with stellar cluster age or mass. For the depth and spatial coverage of these observations, we cannot substantiate the millimeter/submillimeter excess arising from the processing of dust grains by the local interstellar radiation field. We detect a bright unknown source in NGC.628 in ALMA bands 4 and 7 with no counterparts at other wavelengths from ancillary data. We speculate this is possibly a dust-obscured supernova.