Adobe, Inc.

Synopsys, Inc.

Google Inc.

Haichun Chen

Vidit Bhatia

Vijeth Lomada

Huazhong Ning

Ming Su

Scott I Chase

Greg Woolhiser

Techniques for generating new machine learning (ML) systems are described. In an example, a computer system receives a request specifying a task and a performance metric for the new ML model via a user interface. In response, the computer system dynamically generates new feature-extraction rules and new machine learning models based on a rule-model combination that would perform the specified task at a level meeting or exceeding the performance metric.

Generating new machine learning models based on combinations of historical feature-extraction rules and historical machine-learning models

The present disclosure relates to systems, non-transitory computer-readable media, and methods for generating user embeddings utilizing an interaction-to-vector neural network. For example, a user embeddings system transforms unorganized data of user interactions with content items into structured user interaction data. Further, the user embeddings system can utilize the structured user interaction data to train a neural network in a semi-supervised manner and generate uniform vectorized user embeddings for each of the users.

Generating homogenous user embedding representations from heterogeneous user interaction data using a neural network

The present disclosure relates to systems, non-transitory computer-readable media, and methods for expanding user segments automatically utilizing user embedding representations generated by a trained neural network. For example, a user embeddings system expands a segment of users by identifying holistically similar users from uniform user embeddings that encode behavior and/or realized traits of the users. Further, the user embeddings system facilitates the expansion of user segments in a particular direction and focus to improve the accuracy of user segments.

Performing automatic segment expansion of user embeddings using multiple user embedding representation types

The present disclosure relates to systems, non-transitory computer-readable media, and methods for generating user embeddings utilizing an LSTM autoencoder model that captures a history of changes to user trait data. For example, the user embeddings system identifies user trait changes from the user trait data over time as well as generates user trait sequences. Further, the user embeddings system can utilize the user trait sequences to train an LSTM neural network in a semi-supervised manner and generate uniform user embeddings for users.

Generating user embedding representations that capture a history of changes to user trait data

A method is described that includes comprising receiving, by a first computing device and from a second computing device, an indication of user interaction with a first application. The method also includes determining, by the first computing device and based on the user interaction with the first application, a user engagement score associated with the first application, the user engagement score indicating at least one of how frequently or how long the user uses the first application. The method also includes responsive to determining that the user engagement score does not satisfy a threshold user engagement score, determining, by the first computing device and from a set of applications that are alternatives to the first application, a second application to substitute for the first application. The method further includes sending, by the first computing device and to the second computing device, for display, an indication of the second application.

Recommending substitute applications

Some embodiments of the present invention create a layout for a circuit design which includes one or more circuit modules. The system can receive a nominal implementation of a circuit module, and a user-defined module generator capable of generating one or more custom implementations of the circuit module from an existing implementation of the circuit module. Next, the system can create the layout for the circuit design by executing the user-defined module generator on at least one processor to generate one or more custom implementations of the circuit module from the nominal implementation. The system can then use the one or more custom implementations of the circuit module in the layout.

Custom module generation

A computer system and corresponding process is disclosed for making a design layout based on a schematic diagram. The system comprises a user interface which includes a display of a schematic diagram of a layout to be designed. The schematic diagram includes multiple occurrences of a target element. Source elements are displayed, which correspond with the target element. One occurrence of multiple occurrences of the target element shown in the schematic diagram is selected. The source element is applied to the selected target element. A computer program product bears software for directing a computer system to perform the foregoing.

Growing community of inventors

Sunnyvale, CA, United States of America

Haichun Chen