Methods, systems, and devices for event prediction are described. The method may include receiving real-time data associated with one or more measurements by a device. The method may also include updating one or more slope lines based on past instances of the real-time data and generating a feature vector based on the real-time data and the one or more slopes lines. The method may further include passing the first feature vector through an artificial neural network (ANN) and identifying an ANN output that acts as a predictive score. The method may also include predicting whether or not an event will occur based on the predictive score. In some cases, the predictive score may be passed through a voting mechanism, and the prediction of whether or not the event will occur may depend on an output of the voting mechanism.