Abstract: A method of training a reinforcement learning agent to output actions from a continuous action space, comprising: providing an actor ensemble that includes a plurality of actor neural networks that each output a respective action from the continuous action space in response to an observed state of an environment; providing a critic neural network that approximates a state-action value function indicating an impact of an action on the environment based on a reward from the environment and the observed state of the environment; training the actor ensemble and the critic neural network to maximize a state-action value from the state-action value function over successive time steps by, in each time step: selecting from the respective actions output by the plurality of actor neural networks the action that will provide a best state-action value from the state-action value function; applying the selected action to the environment; based on an observed state of the environment of in response to the selected action,

Abstract: Methods and apparatus, including computer program products, are provided for monitoring user interface state transitions. In one aspect, there is provided a method. The method may include monitoring a plurality of operations performed at a user interface; recording, at a cache, the plurality of operations, each of the recorded plurality of operations having a predefined format including a value and a target; moving, in response to a first request to move to a prior state at the user interface, back in the cache to determine the prior state at the user interface; and moving, in response to a second request to move to a subsequent state at the user interface, forward in the cache to determine the subsequent state at the user interface. Related apparatus, systems, methods, and articles are also described.

Abstract: In one embodiment, a method monitors input by a user using an application and detects a first input for a date in the application. A list of candidates is generated for the first input, the list of candidates listing the date in different formats that are possible based on the date. The method uses a space function to retrieve context for the first input, the context selected from information proximate to the first input within a threshold. Also, the method uses a time function to retrieve a history of selections, the history of selections being selected from one or more prior lists of candidates from prior inputs. The list of candidates is ranked based on a ranking function that uses the context and the history of selections to rank the list of candidates. A portion of the ranked list is output on the application for selection by the user.

Abstract: Methods and apparatus, including computer program products, are provided for rendering and re-rendering a user interface of a web application based on definition of and traversal of a minimum sub-metadata tree. Related apparatus, systems, methods, and articles are also described.

Abstract: In one embodiment, a method monitors input by a user using an application and detects a first input for a date in the application. A list of candidates is generated for the first input, the list of candidates listing the date in different formats that are possible based on the date. The method uses a space function to retrieve context for the first input, the context selected from information proximate to the first input within a threshold. Also, the method uses a time function to retrieve a history of selections, the history of selections being selected from one or more prior lists of candidates from prior inputs. The list of candidates is ranked based on a ranking function that uses the context and the history of selections to rank the list of candidates. A portion of the ranked list is output on the application for selection by the user.

Abstract: Methods and apparatus, including computer program products, are provided for rendering and re-rendering a user interface of a web application based on definition of and traversal of a minimum sub-metadata tree. Related apparatus, systems, methods, and articles are also described.

Abstract: Methods and apparatus, including computer program products, are provided for monitoring user interface state transitions. In one aspect, there is provided a method. The method may include monitoring a plurality of operations performed at a user interface; recording, at a cache, the plurality of operations, each of the recorded plurality of operations having a predefined format including a value and a target; moving, in response to a first request to move to a prior state at the user interface, back in the cache to determine the prior state at the user interface; and moving, in response to a second request to move to a subsequent state at the user interface, forward in the cache to determine the subsequent state at the user interface. Related apparatus, systems, methods, and articles are also described.