公开(公告)号：US11093497B1

公开(公告)日：2021-08-17

申请号：US15934240

申请日：2018-03-23

Applicant: Amazon Technologies, Inc.

Inventor： Pracheer Gupta ,  Poorna Chand Srinivas Perumalla ,  Stefano Stefani

IPC: G06F16/24 ,  G06F16/2453 ,  G06F16/22 ,  G06F16/23 ,  G06F16/901

Abstract: Techniques are described for a nearest neighbor search service that enables users to perform nearest neighbor searches. The nearest neighbor search service includes an interface that enables users to create collections of searchable vectors, add and update vectors to a collection, delete vectors from a collection, and perform searches for nearest neighbors to a given vector. The nearest neighbor search service enables users to add, update, and delete vectors of a collection in real-time while also enabling users to perform searches at the same time.

公开(公告)号：US20200236171A1

公开(公告)日：2020-07-23

申请号：US16839821

申请日：2020-04-03

Applicant: Amazon Technologies, Inc.

Inventor： Pracheer Gupta ,  Poorna Chand Srinivas Perumalla ,  Jia Bi Zhang ,  Srikanth Kandalam Srinivasa ,  Madan Mohan Rao Jampani ,  Stefano Stefani

IPC: H04L29/08 ,  G06F16/22 ,  G06F16/27

Abstract: A data set may be partitioned according to relative differences indicated by a cover tree. A cover tree may be generated for a data set. Items in the data set may be stored at the same or different nodes according to the relative difference between the items indicated in the cover tree. Portions of the cover tree may be assigned to different nodes storing the data set. Access requests for the data set may be performed by sending the access requests to nodes identified according to the assigned portions of the cover tree.

公开(公告)号：US10915524B1

公开(公告)日：2021-02-09

申请号：US15634577

申请日：2017-06-27

Applicant: Amazon Technologies, Inc.

Inventor： Pracheer Gupta ,  Poorna Chand Srinivas Perumalla ,  Andrea Olgiati ,  Madan Mohan Rao Jampani ,  Stefano Stefani

IPC: G06F7/00 ,  G06F16/245 ,  G06F16/21 ,  G06F16/27 ,  G06F16/23

Abstract: A computing resource service provider deploys resources to process input data sets on an ongoing basis and provide requestors with queryable data structures generated from the input data sets over determined, rolling periods of time. In one embodiment, the input data sets are processed using one or more nearest neighbor search algorithms, and the outputs therefrom are represented in data structures which are rotated as newer data structures are subsequently generated. The disclosed systems and techniques improve resource utilization, processing efficiency, query latency, and result consistency relative to known controls for large and/or complex data processing tasks, such as those employed in machine learning techniques.

公开(公告)号：US11210605B1

公开(公告)日：2021-12-28

申请号：US15658005

申请日：2017-07-24

Applicant: Amazon Technologies, Inc.

Inventor： Pracheer Gupta ,  Andrea Olgiati ,  Poorna Chand Srinivas Perumalla ,  Stefano Stefani ,  Maden Mohan Rao Jampani

IPC: G06N20/00 ,  G06F17/18 ,  G06F16/28

Abstract: A processing device receives a dataset comprising a plurality of data points, wherein each data point of the plurality of data points comprises a representative vector for the data point and an associated classification for the data point. The processing device determines, for the dataset, a score representative of a degree of clustering of the plurality of data points. The processing device determines a suitability of the dataset for use in machine learning based on the score.

公开(公告)号：US11075991B2

公开(公告)日：2021-07-27

申请号：US16839821

申请日：2020-04-03

Applicant: Amazon Technologies, Inc.

Inventor： Pracheer Gupta ,  Poorna Chand Srinivas Perumalla ,  Jia Bi Zhang ,  Srikanth Kandalam Srinivasa ,  Madan Mohan Rao Jampani ,  Stefano Stefani

IPC: H04L29/08 ,  G06F16/27 ,  G06F16/22

Abstract: A data set may be partitioned according to relative differences indicated by a cover tree. A cover tree may be generated for a data set. Items in the data set may be stored at the same or different nodes according to the relative difference between the items indicated in the cover tree. Portions of the cover tree may be assigned to different nodes storing the data set. Access requests for the data set may be performed by sending the access requests to nodes identified according to the assigned portions of the cover tree.

公开(公告)号：US11055286B2

公开(公告)日：2021-07-06

申请号：US15934277

申请日：2018-03-23

Applicant: Amazon Technologies, Inc.

Inventor： Poorna Chand Srinivas Perumalla ,  Pracheer Gupta ,  Stefano Stefani

IPC: G06F16/24 ,  G06F16/2453 ,  G06F16/22 ,  G06F16/23 ,  G06F16/901

Abstract: Techniques are described for a nearest neighbor search service that enables users to perform nearest neighbor searches. The nearest neighbor search service includes an interface that enables users to create collections of searchable vectors, add and update vectors to a collection, delete vectors from a collection, and perform searches for nearest neighbors to a given vector. The nearest neighbor search service enables users to add, update, and delete vectors of a collection in real-time while also enabling users to perform searches at the same time.

公开(公告)号：US11023440B1

公开(公告)日：2021-06-01

申请号：US15634334

申请日：2017-06-27

Applicant: Amazon Technologies, Inc.

Inventor： Pracheer Gupta ,  Madan Mohan Rao Jampani ,  Andrea Olgiati ,  Poorna Chand Srinivas Perumalla ,  Stefano Stefani

IPC: G06F16/00 ,  G06F16/22 ,  G06F16/901

Abstract: A computing resource service provider deploys resources to process input data sets on an ongoing basis and provide requestors with queryable data structures generated from the input data sets over determined, rolling periods of time. In one embodiment, the input data sets are processed using one or more nearest neighbor search algorithms, and the outputs therefrom are represented in data structures which are rotated as newer data structures are subsequently generated. The disclosed systems and techniques improve resource utilization, processing efficiency, query latency, and result consistency relative to known controls for large and/or complex data processing tasks, such as those employed in machine learning techniques.

公开(公告)号：US10587632B1

公开(公告)日：2020-03-10

申请号：US15718770

申请日：2017-09-28

Applicant: Amazon Technologies, Inc.

Inventor： Poorna Chand Srinivas Perumalla ,  Pracheer Gupta ,  Madan Mohan Rao Jampani

IPC: H04L29/06 ,  G06N3/08

Abstract: A method and system including a neural network configured to detect whether a source of the networks packets is transmitting in accordance with a recognized application protocol. The neural network analyzes a set of network packets to determine a probability that the network pattern corresponds to a network pattern associated with a recognized application protocol. If the probability associated with a first recognized application protocol exceeds a threshold probability value, the transmission of the set of network packets may be classified as being transmitted in accordance with the first recognized application protocol. If the probabilities corresponding to the respective recognized application protocols do not exceed the threshold probability value, the neural network classifies the transmission of the set of network packets as malware.