Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for post-hoc deferral for classification tasks. In particular, a system can perform either post-hoc threshold correction or post-hoc rejector training to account for the cost of deferring model inputs to an expert system for classification.

Abstract: Systems and methods described herein relate to creating or modifying multimedia content using theme-based effects. In accordance with some implementations, a method can comprise the operations of: accessing foundational content; receiving a request to apply a theme to the foundational content; receiving a theme-based effect associated with the theme; receiving a licensing parameter of use associated with the theme; and applying the theme-based effect to the foundational content according to the licensing parameter. Themes can include those created by third-party vendors and those created by one or more users of a method or system implementing features described herein. Additionally, use of different themes can vary in cost and/or use limitations, which can be determined according to licensing parameters associated with the themes.

Abstract: The present invention provides methods for treating and/or preventing inflammatory diseases of the gastrointestinal tract with one or more compounds, or pharmaceutically acceptable salts thereof, disclosed herein, and the use of compositions comprising the same.

Abstract: Systems and methods described herein relate to creating or modifying multimedia content using theme-based effects. In accordance with some implementations, a method can comprise the operations of: accessing foundational content; receiving a request to apply a theme to the foundational content; receiving a theme-based effect associated with the theme; receiving a licensing parameter of use associated with the theme; and applying the theme-based effect to the foundational content according to the licensing parameter. Themes can include those created by third-party vendors and those created by one or more users of a method or system implementing features described herein. Additionally, use of different themes can vary in cost and/or use limitations, which can be determined according to licensing parameters associated with the themes.

Abstract: A method for training a machine learning model, e.g., a neural network, using a regularization scheme is disclosed. The method includes generating regularized partial gradients of losses computed using an objective function for training the machine learning model.

Abstract: The present invention provides methods for treating and/or preventing inflammatory diseases of the gastrointestinal tract with one or more compounds, or pharmaceutically acceptable salts thereof, disclosed herein, and the use of compositions comprising the same.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for accounting for long-tail training data.

Abstract: Described embodiments generally relate to a communications system comprising a first message broker configured to receive a request message from a messaging application executed by a customer computing device, the first message broker being a universal message broker adapted to receive messages from a plurality of different messaging applications; a message processing module configured to receive the request message from the first message broker and to interpret the request message to determine a sub-system of a vendor computing system to which the request message is directed; a first formatter configured to receive the request message from the message processing module, and to format the request message to match the format expected by the determined sub-system; and a second message broker configured to receive the formatted message from the first formatter and to send the formatted message to the determined sub-system.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for multi-stage computationally-efficient inference using a first and second neural network.

Abstract: Techniques for improved analysis and simulation of an IC design are disclosed. Simulation activity for an integrated circuit (IC) design is identified using one or more processors. One or more potential improvements to a simulation of the IC design are generated based on the simulation activity, the one or more potential improvements relating to at least one of signal activity or process activity, during simulation, reflected in the simulation activity. A hardware description language (HDL) design file corresponding to the IC design is modified to indicate the one or more potential improvements to the simulation.

Abstract: Generally, the present disclosure is directed to systems and methods that perform spreadout regularization to enable learning of a multi-class classification model in the federated setting, where each user has access to the positive data associated with only a limited number of classes (e.g., a single class). Examples of such settings include decentralized training of face recognition models or speaker identification models, where in addition to the user specific facial images and voice samples, the class embeddings for the users also constitute sensitive information that cannot be shared with other users.

Abstract: Generally, the present disclosure provides systems and methods for performing machine learning in hyperbolic space. Specifically, techniques are provided which enable the learning of a classifier (e.g., large-margin classifier) for data defined within a hyperbolic space (e.g., which may be particularly beneficial for data that possesses a hierarchical structure).

Abstract: The present invention provides methods for treating and/or preventing inflammatory diseases of the gastrointestinal tract with one or more compounds, or pharmaceutically acceptable salts thereof, disclosed herein, and the use of compositions comprising the same.

Abstract: A method of quantifying resolution times through external networks includes receiving, from an external network, a transmission indicating an incident. The transmission includes a first time at which an indication of the incident was posted to the external network. The method may also include generating an incident record associated with the incident, submitting the incident record to an incident management process to resolve the incident, and receiving an indication that the incident is resolved. The method may additionally include submitting a request to the external network to post the indication that the incident is resolved, and receiving, from the external network, a second time at which the indication that the incident is resolved was posted by the external network.

Abstract: There is provided a method and system for generating a program. The method includes detecting a number of steps for performing a task on a computing device and detecting an example relating to each of the steps, wherein the example includes input data and corresponding output data relating to the step. The method also includes, for each example, determining a rule that transforms the input data to the corresponding output data based on cues including textual features within the input data and the corresponding output data. The method further includes generating a program for performing the task based on the rules.

Abstract: The present disclosure provides methods of preparing carbocyclic nucleosides, in particular, Prurisol™ ((?) cis-[4-[2-amino-6-(cyclopropylamino)-9H-purin-9-yl]-2-cyclopentene-1-hydroxymethyl acetate).

Abstract: A method of quantifying resolution times through external networks includes receiving, from an external network, a transmission indicating an incident. The transmission includes a first time at which an indication of the incident was posted to the external network. The method may also include generating an incident record associated with the incident, submitting the incident record to an incident management process to resolve the incident, and receiving an indication that the incident is resolved. The method may additionally include submitting a request to the external network to post the indication that the incident is resolved, and receiving, from the external network, a second time at which the indication that the incident is resolved was posted by the external network.

Abstract: Disclosed are nitrile derivatives and pharmaceutical compositions comprising nitrile derivatives. The pharmaceutical compositions comprise compounds of the formula I and the pharmaceutically acceptable salts of such compounds. Also disclosed are processes for the preparation of such compounds, intermediates used in the preparation of such compounds, and the uses of such compounds in treating hyperproliferative diseases, inflammatory diseases and viral and bacterial infections and inducing apoptosis in cancer cells.

Abstract: Systems and methods described herein relate to creating or modifying multimedia content using theme-based effects. In accordance with some implementations, a method can comprise the operations of: accessing foundational content having an associated content timeline; receiving a request to apply a theme to the foundational content, wherein the theme includes one or more theme-based effects that relate to the theme; receiving a theme-based effect associated with the theme, wherein the theme-based effect has an associated effect timeline; and applying the theme-based effect to the foundational content according to a set of cues associated with the associated content timeline, wherein applying the theme-based effect comprises adapting the associated effect timeline according to the set of cues while preserving the associated content timeline.

Abstract: There is provided a method and system for generating a program. The method includes detecting a number of steps for performing a task on a computing device and detecting an example relating to each of the steps, wherein the example includes input data and corresponding output data relating to the step. The method also includes, for each example, determining a rule that transforms the input data to the corresponding output data based on cues including textual features within the input data and the corresponding output data. The method further includes generating a program for performing the task based on the rules.