Microsoft Technology Licensing, LLC

Todd Douglas Mytkowicz

Madanlal S Musuvathi

Saeed Maleki

Yufei Ding

Aman Kansal

Benjamin Livshits

Deepak Ganesan

Kathryn Stuart McKinley

Mohamed Musthag

Neelakantan Sundaresan

Kristin E Lauter

Alexey Svyatkovskiy

Hao Chen

Shuvendu K Lahiri

Christian Alma Bird

Kim Laine

Mark Marron

Veselin Raychev

Kathryn S Mckinley

Adrian Sampson

Hitesh Kanwathirtha

Arunesh Chandra

Olli Ilari Saarikivi

Georgios Kastrinis

Roshan Dathathri

Christian Bird

Elizabeth Dinella

Olli Llari Saarikivi

An automated system for resolving program merges uses neural transformers with attention. In one aspect, a neural encoder transformer model is trained from developer-resolved merge conflicts to learn to predict a resolution strategy that aids a developer in constructing a merged program. In a second aspect, a neural decoder transformer model is trained on the syntax and semantics of different source code programming languages to predict a merge resolution consisting of interleaved lines of source code from programs A, B, or O, where programs A and B contain changes to code base O.

Automated merge conflict resolution with transformers

Systems, methods, and computer-readable media are disclosed for parallel stochastic gradient descent using linear and non-linear activation functions. One method includes: receiving a set of input examples; receiving a global model; and learning a new global model based on the global model and the set of input examples by iteratively performing the following steps: computing a plurality of local models having a plurality of model parameters based on the global model and at least a portion of the set of input examples; computing, for each local model, a corresponding model combiner based on the global model and at least a portion of the set of input examples; and combining the plurality of local models into the new global model based on the current global model and the plurality of corresponding model combiners.

Systems, methods, and computer-readable media for parallel stochastic gradient descent with linear and non-linear activation functions

Embodiments of the present invention provide systems, methods, and computer storage media directed to optimizing the generation, evaluation, and selection of tensor circuit specifications for a tensor circuit to perform homomorphic encryption operations on encrypted data. A computing device having an improved compiler and runtime configuration can obtain a tensor circuit and associated schema. The computing device can map the obtained tensor circuit to an equivalent tensor circuit, adapted to perform fully homomorphic encryption (FHE) operations, and instantiated based on the obtained associated scheme. The computing device can then monitor a flow of data through the equivalent FHE-adapted tensor circuit utilizing various tensor circuit specifications determined therefor. A cost of each tensor circuit specification can be determined by the computing device based on the monitored flow of data, so as to identify an optimal set of optimal tensor circuit specifications that can be employed by the obtained tensor circuit, to efficiently perform homomorphic encryption operations on encrypted data.

Homomorphic evaluation of tensor programs

Described herein is a system that transmits and combines local models, that individually comprise a set of local parameters computed via stochastic gradient descent (SGD), into a global model that comprises a set of global model parameters. The local models are computed in parallel at different geographic locations along with symbolic representations. The symbolic representations can be used to combine the local models. The global model can determine a likelihood, given a new data instance of a feature set, that a user performs a computer interaction with the content element. For instance, the system can use the model to provide search results in response to a search query submitted by a user. Or, the system can use the model to make a recommendation or suggestion to a user in response to a request for content (e.g., display a targeted advertisement, suggest a news story, etc.).

Determining a likelihood of a user interaction with a content element

Various technologies described herein pertain to detection of an opportune time period to deliver a notification. Responsive to receipt of the notification (e.g., at a user device), analysis of an attention state of a user can be initialized. Further, the opportune time period to deliver the notification can be detected based on the analysis of the attention state of the user. The opportune time period can be during a breakpoint or an influential context. The breakpoint is when the user has switched between tasks and lacks engagement with the tasks. The influential context is a particular context in which the user is available to attend to the notification. Moreover, the notification can be delivered during the opportune time period.

User interruptibility aware notifications

Described herein is a system that transmits and combines local models, that individually comprise a set of local parameters computed via stochastic gradient descent (SGD), into a global model that comprises a set of global model parameters. The local models are computed in parallel at different geographic locations along with symbolic representations. Network transmission of the local models and the symbolic representations, rather than transmission of the large training data subsets processed to compute the local models and symbolic representations, conserves resources and decreases latency. The global model can then be used as a model to determine a likelihood of a course of action being successful for an organization. For example, the course of action can be a purchase of a security or a business operation strategy. In another example, the course of action can be a type of medical treatment for a patient.

Determining a course of action based on aggregated data

Systems, methods, and computer media for machine learning through a symbolic, parallelized stochastic gradient descent (SGD) analysis are provided. An initial data portion analyzer can be configured to perform, using a first processor, SGD analysis on an initial portion of a training dataset. Values for output model weights for the initial portion are initialized to concrete values. Local model builders can be configured to perform, using an additional processor for each local model builder, symbolic SGD analysis on an additional portion of the training dataset. The symbolic SGD analysis uses a symbolic representation as an initial state for output model weights for the corresponding portions of the training dataset. The symbolic representation allows the SGD analysis and symbolic SGD analysis to be performed in parallel. A global model builder can be configured to combine outputs of the local model builders and the initial data portion analyzer into a global model.

Machine learning through parallelized stochastic gradient descent

Described herein is a system transmits and combines local models, that individually include a set of local parameters computed via stochastic gradient descent (SGD), into a global model that includes a set of global model parameters. The local models are computed in parallel at different geographic locations (e.g., different instances of computing infrastructure) along with symbolic representations. Network transmission of the local models and the symbolic representations, rather than transmission of the large training data subsets processed to compute the local models and symbolic representations, conserves resources and decreases latency. The global model can then be used as a model to determine a likelihood that at least a portion of current and/or recently received data traffic is illegitimate data traffic that is associated with a cyber attack. In some instances, the system can implement a remedial action to mitigate the effects of the cyber attack on computing infrastructure.

Implementing network security measures in response to a detected cyber attack

Various technologies described herein pertain to automation of crowd-sourced polling. At least one query can be received. The at least one query includes a request. A poll can be automatically generated based upon the at least one query, where the poll corresponds to the request. The poll can be submitted to a crowdsourcing backend, where instances of the poll are administered on the crowdsourcing backend. Moreover, crowd-sourced responses to the instances of the poll can be retrieved from the crowdsourcing backend. The crowd-sourced responses to the instances of the poll can respectively include crowd-sourced responses to the request. The crowd-sourced responses to the request can be converted to a random variable. An operation can be performed upon the random variable. The operation can include one or more of a statistical analysis (e.g., hypothesis testing), bias correction, an arithmetic operation, expected value computation, standard deviation computation, etc.

Automation of crowd-sourced polling

Described herein is a system that transmits and combines local models, that individually comprise a set of local parameters computed via stochastic gradient descent (SGD), into a global model that comprises a set of global model parameters. The local models are computed in parallel at different geographic locations along with symbolic representations. Network transmission of the local models and the symbolic representations, rather than transmission of the large training data subsets processed to compute the local models and symbolic representations, conserves resources and decreases latency. The global model can then be used as a model to determine a likelihood of a monitored resource or a user of the monitored resource experiencing a problem with respect to performance or completion of one or more operations. The system can also implement an action to assist in resolving or avoiding the problem.

Determining a likelihood of a resource experiencing a problem based on telemetry data

Methods and systems for performance measurements of a program are provided. An execution trace of the program may be captured and stored. The stored execution trace may be replayed in an offline mode. Performance measurements for the program may be determined based on the replaying of the execution trace in the offline mode.

Performance diagnostics via offline replay

A 'Poll Optimizer' provides automated techniques for performing various combinations of both static and runtime optimizations for crowd-sourced queries including, but not limited to, crowd-sourced opinion-based polls. These optimizations have been observed to improve poll performance by reducing factors such as completion times, monetary costs, and error rates of polls. In various implementations, the Poll Optimizer receives an input query representing a crowd-sourced poll that is formatted as a multi-layer structure (e.g., LINQ-based queries natively supported by .NET languages, JQL-based queries supported by JAVA, etc.). The Poll optimizer then iteratively reduces the multi-layer structure of the input query to construct a reformulated query. This reformulated query is then matched to an optimized execution process selected from a plurality of predefined execution processes. Finally, the reformulated query and matching optimized execution process are provided as an optimized version of the crowd-sourced poll for execution via a computer-based crowd-sourcing backend.

Optimizing efficiency and cost of crowd-sourced polling

Concepts and technologies are described herein providing technologies for mining patterns in temporal data streams. Data is broken into data sub-portions. Dependencies in computation between one or more of the data sub-portions are broken using symbolic data types. Symbolic summaries of computations of sub-portions are performed in parallel and are reduced to generate an output.

Technologies for mining temporal patterns in big data

Various techniques for evaluating probabilistic assertions are described herein. In one example, a method includes transforming a program, a probabilistic assertion, and an input into an intermediate representation, the intermediate representation including a Bayesian network of nodes representing distributions. The method further includes verifying a probabilistic assertion in the program using the intermediate representation.

Probabilistic assertions and verifying them

Various technologies described herein pertain to parallel local sequence alignment that aligns a query sequence with a database sequence. The database sequence is segmented into a plurality of stripes. A first processing unit can compute Smith-Waterman values for a first stripe of the database sequence across the query sequence based on a cost function that models biological similarity between sequences. Moreover, a second processing unit can compute Smith-Waterman values for a second stripe of the database sequence across the query sequence based on the cost function. Further, a subset of the Smith-Waterman values for the second stripe of the database sequence across the query sequence can be re-computed based on the cost function (e.g., by the first processing unit or the second processing unit). The subset of the Smith-Waterman values to be re-computed can be determined based on a query sequence length and the cost function.

Growing community of inventors

Redmond, WA, United States of America

Todd Douglas Mytkowicz