In one embodiment, a method includes accessing time series data of a wireless communication link between a first node and a second node on a millimeter-wave communications network; determining that an anomaly exists in the time series data; determining an anomaly type for the anomaly by inputting one or more features derived from the time series data into a machine learning module; generating a report comprising an indication that the anomaly exists and a description of the anomaly type, and determining one or more treatments for the determined anomaly.