Amazon Technologies, Inc.

Donghui Zhuo

Sriram Krishnan

Aleksandr Mikhaylovich Ingerman

Nicolle M Correa

Robert Matthias Steele

Saman Zarandioon

Charles Eric Dannaker

Sudhakar Rao Puvvadi

Tarun Agarwal

Daniel Blick

Jun Qian

Polly Po Lee

Ralf Herbrich

Jin Li

Michael Brueckner

Jon Arron McClintock

Brandon William Porter

Leo Parker Dirac

Tianming Zheng

Melissa Elaine Davis

Anton Vladilenovich Goldberg

Aram Grigoryan

Rakesh Ramakrishnan

Polly Po Yee Lee

Matthew Paul Wenger

Consistency metadata, including a parameter for a pseudo-random number source, are determined for training-and-evaluation iterations of a machine learning model. Using the metadata, a first training set comprising records of at least a first chunk is identified from a plurality of chunks of a data set. The first training set is used to train a machine learning model during a first training-and-evaluation iteration. A first test set comprising records of at least a second chunk is identified using the metadata, and is used to evaluate the model during the first training-and-evaluation iteration.

Consistent filtering of machine learning data

Pattern based detection of data usage is facilitated using data injection. Data values are injected in one or more storage locations accessible to a plurality of services or included in service requests. Service interactions among the services are compared to a set of patterns. The set of patterns are configured to match the data values. By comparing the service interactions to the patterns, one or more of the service interactions are determined to include individual ones of the data values. Data are generated indicating a presence of the data values in the services.

Pattern-based detection using data injection

A record extraction request for a data set is received at a machine learning service. A plan to perform one or more chunk-level operations (such as sampling, shuffling, splitting or partitioning for parallel computation) on chunks of the data set is generated. A set of data transfers that results in a particular chunk being stored in a particular server's memory is initiated to implement the first chunk-level operation of the sequence. A second operation such as another filtering operation or a feature processing operation is performed on a result set of the first chunk-level operation.

Input processing for machine learning

Methods, systems, and computer-readable media for implementing pattern-based detection are disclosed. A plurality of services monitor a plurality of service interactions comprising data or metadata. The services compare the data or metadata to a set of patterns and identify one or more matched patterns among the set of patterns. The services send data indicative of the matched patterns to a central recording service. The central recording service aggregates the data indicative of the matched patterns and generates one or more data flow visualizations indicating one or more data flows between individual ones of the services for the matched patterns.

Pattern-based detection for services in distributed systems

Growing community of inventors

Kirkland, WA, United States of America

Donghui Zhuo