retrieving, by the one or more computers, data indicating labels for clusters of network performance anomalies, wherein the clusters respectively correspond to different network conditions experienced by satellite terminals of a satellite communication system, wherein the clusters each include multiple network performance anomalies grouped according to frequency of co-occurrence such that the clusters respectively represent different groups of network performance anomalies that are indicated by a data set to co-occur for individual satellite terminals of the satellite communication system, and wherein each label for a cluster indicates a network condition classification for a network condition associated with the network performance anomalies in the cluster;

generating, by the one or more computers, a set of training data to train a machine learning model, the set of training data being generated by assigning the labels for the clusters to sets of performance indicators used to generate the clusters, the labels being assigned based on levels of commonality among types of network performance anomalies indicated by respective sets of performance indicators and the types of network performance anomalies included in the respective clusters;

training, by the one or more computers, a machine learning model to predict network condition classifications for satellite terminals based on input of performance indicators for the satellite terminals, wherein the machine learning model is trained to evaluate satellite terminals with respect to a predetermined set of network condition classifications including the network condition classifications for the clusters, and wherein the machine learning model is trained using training data examples from the set of training data in which each training example includes a set of performance indicators and the label assigned for the set of performance indicators; and

for each of one or more satellite terminals, determining, by the one or more computers, a network condition classification for the satellite terminal based on output that the trained machine learning model generates based on input of performance indicators for the satellite terminal.