Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a neural network that is configured to process an input image to generate a network output for the input image. In one aspect, a method comprises, at each of a plurality of training steps: obtaining a plurality of training images for the training step; obtaining, for each of the plurality of training images, a respective target output; and selecting, from a plurality of image patch generation schemes, an image patch generation scheme for the training step, wherein, given an input image, each of the plurality of image patch generation schemes generates a different number of patches of the input image, and wherein each patch comprises a respective subset of the pixels of the input image.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network using re-initialization. One of the methods includes, at each time step in a sequence of time steps: identifying current values of the weights as of the training time step; selecting one of the layer blocks; generating new values for the weights of the plurality of neural network layers, comprising: re-initializing the values of the weights of at least the neural network layers in the layer blocks that are after the selected layer block without re-initializing the current values of the weights of the neural network layers in the layer block and the neural network layers in any layer block that is before the selected layer block; and raining the neural network starting from the new values for the weights of the plurality of neural network layers.

Abstract: Systems and methods for managing deliveries to a hydrocarbon storage system that includes a plurality of hydrocarbon storage facilities include a first machine-learning model for each individual hydrocarbon storage facility that predicts truck-waiting times and sales volumes and a second machine-learning model for the hydrocarbon storage system that outputs a recommended hauling volume for each individual hydrocarbon storage facility.

Abstract: Systems and methods include a method for optimizing an action at a facility using a prediction of a target variable. Historical data is collected for a set of facilities. The historical data includes transactional data for discrete events that occurred at the set of facilities and non-transactional data spanning continuous time periods. Semidefinite matrices are generated using the historical data, where the semidefinite matrices incorporate historical samples of a form (yt, xt, zt), and where yt is a target variable to be predicted at time t, xt is an input parameter at time t, and zt is an environment variable at time t. A statistical model based on the semidefinite matrices is determined. Production at a facility is monitored, including collecting production data comprising the transactional data and the non-transactional data of the facility. Using the statistical model and the production data, a prediction of a target variable associated with operating conditions at the facility is determined.