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.Plain Language Summary Seismic stations record not only earthquake signals but also a wide variety of nuisance signals. Some of these nuisance signals are impulsive and can initially look very similar to real earthquake signals. This is a problem for earthquake early warning (EEW) algorithms, which sometimes misinterpret such signals as being real earthquake signals, and which may then send out false alerts. For each registered impulsive signal, EEW systems need to decide (or, classify) in real-time whether or not the signal stems from an actual earthquake. State-of-the-art machine learning (ML) classifiers have been shown to strongly outperform more standard linear classifiers in a wide range of classification problems. Here we analyze the performance of a variety of different ML classifiers to identify which type of classifier leads to the most reliable signal/noise discrimination in an EEW context. We find that we can successfully train complex deep learning classifiers that can discriminate between nuisance and earthquake signals very reliably (accuracy of 99.5%). Less complex ML classifiers also outperform a linear classifier, but with significantly higher error rates. The deep ML classifiers may allow EEW systems to almost entirely avoid false and missed signal detections...