Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for controlling agents. In particular, an agent can be controlled using a discretization neural network that generates a state-dependent discretization of an original action space and a policy neural network that is used to select an action from the state-dependent quantization rather than from the original action space.

Abstract: A method is provided for characterising a gait comprising steps of: receiving a signal measuring the movement of a target subject measured beforehand; comparing (204) at least one section of the measuring signal with a plurality of reference signals, the plurality of reference signals comprising at least one movement signal measured beforehand on a reference subject different from the target subject, each comparison of the measuring signal with a reference signal producing a similarity coefficient representative of a probability of similarity between the compared signals; and selecting (208) at least one local maximum of interest among the similarity coefficients produced, wherein the plurality of reference signals comprises at least one movement signal measured beforehand on a reference subject afflicted with a gait disorder, and at least one movement signal measured beforehand on a healthy reference object.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network used to select actions performed by an agent interacting with an environment by performing actions that cause the environment to transition states. One of the methods includes obtaining a transition generated as a result of the reinforcement learning agent interacting with the environment, processing a bonus input using a bonus estimation neural network to generate an exploration bonus estimate that encourages the agent to explore the environment in accordance with an expert exploration strategy that would be adopted by an expert agent; generating a modified reward from the reward included in the transition and the exploration bonus estimate; and determining an update to current parameter values of the neural network to optimize a reinforcement learning objective function that maximizes returns to be received by the agent with respect to the modified reward.

Abstract: A method is provided for characterising a gait comprising steps of: receiving a signal measuring the movement of a target subject measured beforehand; comparing (204) at least one section of the measuring signal with a plurality of reference signals, the plurality of reference signals comprising at least one movement signal measured beforehand on a reference subject different from the target subject, each comparison of the measuring signal with a reference signal producing a similarity coefficient representative of a probability of similarity between the compared signals; and selecting (208) at least one local maximum of interest among the similarity coefficients produced, wherein the plurality of reference signals comprises at least one movement signal measured beforehand on a reference subject afflicted with a gait disorder, and at least one movement signal measured beforehand on a healthy reference object.