In earthquake early warning (EEW), every sufficiently impulsive signal is potentially the first evidence for an unfolding large earthquake. More often than not, however, impulsive signals are mere nuisance signals. One of the most fundamental—and difficult—tasks in EEW is to rapidly and reliably discriminate real local earthquake signals from all other signals. This discrimination is necessarily based on very little information, typically a few seconds worth of seismic waveforms from a small number of stations. As a result, current EEW systems struggle to avoid discrimination errors and suffer from false and missed alerts. In this study we show how modern machine learning classifiers can strongly improve real‐time signal/noise discrimination. We develop and compare a series of nonlinear classifiers with variable architecture depths, including fully connected, convolutional and recurrent neural networks, and a model that combines a generative adversarial network with a random forest. We train all classifiers on the same data set, which includes 374 k local earthquake records (M3.0–9.1) and 946 k impulsive noise signals. We find that all classifiers outperform existing simple linear classifiers and that complex models trained directly on the raw signals yield the greatest degree of improvement. Using 3‐s‐long waveform snippets, the convolutional neural network and the generative adversarial network with a random forest classifiers both reach 99.5% precision and 99.3% recall on an independent validation data set. Most misclassifications stem from impulsive teleseismic records, and from incorrectly labeled records in the data set. Our results suggest that machine learning classifiers can strongly improve the reliability and speed of EEW alerts.