Galileo is a software solution that performs unsupervised Machine Learning to target adversary frequency-hopping tactical radios, isolate frequencies and hop patterns, and perform automatic predictive EW responses. Galileo uses our advanced algorithms to blindly detect, characterize, and infer EW targeting parameters with no static target library. Galileo labels known threats in near real-time, continuously providing updated training sets to its detectors. Galileo not only detects fast frequency hopping signals with no prior information, but also identifies small observed variances in known threat emitters. It detects individual threat hops, clusters multiple emitters simultaneously, and in most cases automatically demodulates and frame-synchronizes anomalous emitters.