An embodiment of a spoken intent detection device includes technology to detect a phrase in an electronic representation of an audio stream based on a pre-defined vocabulary, associate a time stamp with the detected phrase, and classify a spoken intent based on a sequence of detected phrases and the respective associated time stamps. Other embodiments are disclosed and claimed.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating neural network architectures. One of the methods includes receiving a request to determine an architecture for a task neural network; maintaining data specifying a plurality of candidate architectures for the task neural network; repeatedly performing operations comprising: selecting one or more candidate architectures in the maintained data to be modified; generating a new candidate architecture from the selected candidate architecture by, for each hyperparameter in the set of hyperparameters, selecting the value for the hyperparameter for the new candidate architecture; and adding data specifying the new candidate architecture to the maintained data; and selecting, as the final architecture for the task neural network, one of the candidate architectures specified in the maintained data.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating neural network architectures. One of the methods includes receiving a request to determine an architecture for a task neural network; maintaining data specifying a plurality of candidate architectures for the task neural network; repeatedly performing operations comprising: selecting one or more candidate architectures in the maintained data to be modified; generating a new candidate architecture from the selected candidate architecture by, for each hyperparameter in the set of hyperparameters, selecting the value for the hyperparameter for the new candidate architecture; and adding data specifying the new candidate architecture to the maintained data; and selecting, as the final architecture for the task neural network, one of the candidate architectures specified in the maintained data.

A method includes generating an intervention model for a population of users based on contact data, demographic data, and engagement data indicating successfulness of prior interventions for each of the population of users. The method includes, obtaining first data related to a first user, including engagement data indicating successfulness of prior interventions with the first user. The method includes supplying the obtained data as input to the intervention model to determine an intervention expectation, which indicates a likelihood that the first user will take action in response to an intervention. The method includes determining a likelihood of a gap in care. The method includes, in response to the care gap likelihood exceeding a minimum threshold, selecting and scheduling execution of a first intervention. The first intervention is one of a real-time communication with the first user by a specialist and an automated transmission of a message to the first user.

A method includes generating an intervention model for a population of users based on contact data, demographic data, and engagement data indicating successfulness of prior interventions for each of the population of users. The method includes, obtaining first data related to a first user, including engagement data indicating successfulness of prior interventions with the first user. The method includes supplying the obtained data as input to the intervention model to determine an intervention expectation, which indicates a likelihood that the first user will take action in response to an intervention. The method includes determining a likelihood of a gap in care. The method includes, in response to the care gap likelihood exceeding a minimum threshold, selecting and scheduling execution of a first intervention. The first intervention is one of a real-time communication with the first user by a specialist and an automated transmission of a message to the first user.

A modulated segmentation system can use a modulator network to emphasize spatial prior data of an object to track the object across multiple images. The modulated segmentation system can use a segmentation network that receives spatial prior data as intermediate data that improves segmentation accuracy. The segmentation network can further receive visual guide information from a visual guide network to increase tracking accuracy via segmentation.

A modulated segmentation system can use a modulator network to emphasize spatial prior data of an object to track the object across multiple images. The modulated segmentation system can use a segmentation network that receives spatial prior data as intermediate data that improves segmentation accuracy. The segmentation network can further receive visual guide information from a visual guide network to increase tracking accuracy via segmentation.

In a general aspect, input data for a computer process are preprocessed by a preprocessor unit that includes a quantum processor. In some aspects, a preprocessor unit obtains input data for a computer process that is configured to run on a computer processing unit. Randomized parameter values are computed for variable parameters of a quantum logic circuit based on the input data. A classical processor in the preprocessor unit computes the randomized parameter values from the input data and a set of random numbers. A quantum processor in the preprocessor unit produces quantum processor output data by executing the quantum logic circuit having the randomized parameter values assigned to the variable parameters. Preprocessed data generated based on the quantum processor output data are then provided as the input for the computer process configured to run on the computer processing unit.

In a general aspect, input data for a computer process are preprocessed by a preprocessor unit that includes a quantum processor. In some aspects, a preprocessor unit obtains input data for a computer process that is configured to run on a computer processing unit. Randomized parameter values are computed for variable parameters of a quantum logic circuit based on the input data. A classical processor in the preprocessor unit computes the randomized parameter values from the input data and a set of random numbers. A quantum processor in the preprocessor unit produces quantum processor output data by executing the quantum logic circuit having the randomized parameter values assigned to the variable parameters. Preprocessed data generated based on the quantum processor output data are then provided as the input for the computer process configured to run on the computer processing unit.

A computer implemented method of improving parameters of a critic approximator module includes receiving, by a mixed integer program (MIP) actor, (i) a current state and (ii) a predicted performance of an environment from the critic approximator module. The MIP actor solves a mixed integer mathematical problem based on the received current state and the predicted performance of the environment. The MIP actor selects an action a and applies the action to the environment based on the solved mixed integer mathematical problem. A long-term reward is determined and compared to the predicted performance of the environment by the critic approximator module. The parameters of the critic approximator module are iteratively updated based on an error between the determined long-term reward and the predicted performance.