公开(公告)号：US20160239736A1

公开(公告)日：2016-08-18

申请号：US14624500

申请日：2015-02-17

Applicant: QUALCOMM Incorporated

Inventor： Anthony SARAH

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06N3/04 ,  G06N3/082

Abstract: A method for configuring a classifier includes operating the classifier to classify an input. The method also includes determining a confidence metric based on classification of the input. The method further includes dynamically updating a complexity of the classifier based on the confidence metric. The confidence metric may be computed based on a posterior probability. The complexity may be updated when the confidence metric is below a threshold value.

Abstract translation: 用于配置分类器的方法包括操作分类器以对输入进行分类。 该方法还包括基于输入的分类来确定置信度量度。 该方法还包括基于置信度度动态地更新分类器的复杂性。 可以基于后验概率来计算置信量度。 当置信量度低于阈值时，可以更新复杂度。

公开(公告)号：US20170039469A1

公开(公告)日：2017-02-09

申请号：US14849518

申请日：2015-09-09

Applicant: QUALCOMM Incorporated

Inventor： Somdeb MAJUMDAR ,  Dexu LIN ,  Regan Blythe TOWAL ,  Anthony SARAH

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06K9/4628 ,  G06K9/6256 ,  G06K9/6269 ,  G06N3/0445 ,  G06N3/0454 ,  G06N20/00

Abstract: A method of detecting unknown classes is presented and includes generating a first classifier for multiple first classes. In one configuration, an output of the first classifier has a dimension of at least two. The method also includes designing a second classifier to receive the output of the first classifier to decide whether input data belongs to the multiple first classes or at least one second class.

Abstract translation: 提出了一种检测未知类的方法，包括生成多个第一类的第一分类器。 在一个配置中，第一分类器的输出具有至少两个维度。 该方法还包括设计第二分类器以接收第一分类器的输出以确定输入数据是属于多个第一类还是至少一个第二类。

公开(公告)号：US20190228311A1

公开(公告)日：2019-07-25

申请号：US15877723

申请日：2018-01-23

Applicant: QUALCOMM Incorporated

Inventor： Anthony SARAH ,  Raghuraman KRISHNAMOORTHI

IPC: G06N3/08 ,  G06N3/04

Abstract: An apparatus of operating a computational network is configured to determine a low-rank approximation for one or more layers of the computational network based at least in part on a set of residual targets. A set of candidate rank vectors corresponding to the set of residual targets may be determined. Each of the candidate rank vectors may be evaluated using an objective function. A candidate rank vector may be selected and used to determine the low rank approximation. The computational network may be compressed based on the low-rank approximation. In turn the computational network may be operated using the one or more compressed layers.

公开(公告)号：US20160328644A1

公开(公告)日：2016-11-10

申请号：US14878689

申请日：2015-10-08

Applicant: QUALCOMM Incorporated

Inventor： Dexu LIN ,  Venkata Sreekanta Reddy ANNAPUREDDY ,  Sachin Subhash TALATHI ,  Mark STASKAUSKAS ,  Aniket VARTAK ,  Regan Blythe TOWAL ,  David Jonathan JULIAN ,  Anthony SARAH

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06K9/4628 ,  G06N3/04 ,  G06N3/0454 ,  G06N3/084

Abstract: A method of adaptively selecting a configuration for a machine learning process includes determining current system resources and performance specifications of a current system. A new configuration for the machine learning process is determined based at least in part on the current system resources and the performance specifications. The method also includes dynamically selecting between a current configuration and the new configuration based at least in part on the current system resources and the performance specifications.

Abstract translation: 自动选择用于机器学习过程的配置的方法包括确定当前系统的当前系统资源和性能规范。 至少部分地基于当前系统资源和性能规格确定用于机器学习过程的新配置。 该方法还包括至少部分地基于当前系统资源和性能规范来动态地在当前配置和新配置之间进行选择。

公开(公告)号：US20160328645A1

公开(公告)日：2016-11-10

申请号：US14882351

申请日：2015-10-13

Applicant: QUALCOMM Incorporated

Inventor： Dexu LIN ,  Matthew BADIN ,  David Edward HOWARD ,  Daniel Hendricus Franciscus DIJKMAN ,  Michael Colin TREMAINE ,  Anthony SARAH

IPC: G06N3/08 ,  G06N99/00

CPC classification number: G06N3/08 ,  G06N3/063 ,  G06N20/00

Abstract: A method of reducing computational complexity for a fixed point neural network operating in a system having a limited bit width in a multiplier-accumulator (MAC) includes reducing a number of bit shift operations when computing activations in the fixed point neural network. The method also includes balancing an amount of quantization error and an overflow error when computing activations in the fixed point neural network.

Abstract translation: 一种降低在乘法器 - 累加器（MAC）中具有有限位宽度的系统中操作的固定点神经网络的计算复杂度的方法包括在计算固定点神经网络中的激活时减少多个位移操作。 该方法还包括在计算固定点神经网络中的激活时平衡量化误差量和溢出误差。

公开(公告)号：US20150324689A1

公开(公告)日：2015-11-12

申请号：US14483075

申请日：2014-09-10

Applicant: QUALCOMM Incorporated

Inventor： Casimir Matthew WIERZYNSKI ,  David Jonathan JULIAN ,  Anthony SARAH

IPC: G06N3/08

CPC classification number: G06N3/049 ,  G06N3/0454 ,  G06N3/08

Abstract: A method of updating a set of classifiers includes applying a first set of classifiers to a first set of data. The method further includes requesting, from a remote device, a classifier update based on an output of the first set of classifiers or a performance measure of the application of the first set of classifiers.

Abstract translation: 更新一组分类器的方法包括将第一组分类器应用于第一组数据。 该方法还包括从远程设备请求基于第一组分类器的输出的分类器更新或第一组分类器的应用的性能测量。

公开(公告)号：US20150112908A1

公开(公告)日：2015-04-23

申请号：US14056856

申请日：2013-10-17

Applicant: QUALCOMM Incorporated

Inventor： Anthony SARAH ,  Robert Howard KIMBALL ,  Brian SPINAR

IPC: G06N3/08

CPC classification number: G06N3/08 ,  G06N3/049

Abstract: A method for dynamically modifying synaptic delays in a neural network includes initializing a delay parameter and operating the neural network. The method further includes dynamically updating the delay parameter based on a program which is based on a statement including the delay parameter.

Abstract translation: 用于动态修改神经网络中的突触延迟的方法包括初始化延迟参数并操作神经网络。 该方法还包括基于基于包括延迟参数的语句的程序来动态地更新延迟参数。

公开(公告)号：US20150100529A1

公开(公告)日：2015-04-09

申请号：US14085740

申请日：2013-11-20

Applicant: QUALCOMM INCORPORATED

Inventor： Anthony SARAH ,  Robert Howard KIMBALL ,  Michael-David Nakayoshi CANOY ,  Jan Krzys WEGRZYN

IPC: G06N3/08

CPC classification number: G06N3/10 ,  G06N3/049

Abstract: A method of generating executable code for a target platform in a neural network includes receiving a spiking neural network description. The method also includes receiving platform-specific instructions for one or more target platforms. Further, the method includes, generating executable code for the target platform(s) based on the platform-specific instructions and the network description.

Abstract translation: 在神经网络中为目标平台生成可执行代码的方法包括接收尖峰神经网络描述。 该方法还包括接收针对一个或多个目标平台的平台特定指令。 此外，该方法包括：基于特定于平台的指令和网络描述，为目标平台生成可执行代码。

公开(公告)号：US20150095273A1

公开(公告)日：2015-04-02

申请号：US14066599

申请日：2013-10-29

Applicant: QUALCOMM Incorporated

Inventor： Anthony SARAH ,  Robert Howard KIMBALL

IPC: G06N3/02 ,  G06N3/08

CPC classification number: G06N3/049 ,  G06N3/04 ,  G06N3/063 ,  G06N3/08 ,  G06N3/082 ,  G06N99/005

Abstract: A method for improving neural dynamics includes obtaining prototypical neuron dynamics. The method also includes modifying parameters of a neuron model so that the neuron model matches the prototypical neuron dynamics. The neuron dynamics comprise membrane voltages and/or spike timing.

Abstract translation: 改进神经动力学的方法包括获得原型神经元动力学。 该方法还包括修改神经元模型的参数，使得神经元模型与原型神经元动力学匹配。 神经元动力学包括膜电压和/或尖峰定时。

公开(公告)号：US20140310216A1

公开(公告)日：2014-10-16

申请号：US14045672

申请日：2013-10-03

Applicant: QUALCOMM Incorporated

Inventor： Anthony SARAH ,  Jeffrey Alexander LEVIN

IPC: G06N3/02

CPC classification number: G06N3/10 ,  G06N3/049

Abstract: A method generates compact representations of spike timing-dependent plasticity (STDP) curves. The method includes segmenting a set of data points into different sections. The method further includes representing at least one section as a primitive and storing parameters of the primitive. The primitive can be a polynomial.

Abstract translation: 一种方法产生尖峰时间依赖可塑性（STDP）曲线的紧凑表示。 该方法包括将一组数据点分割成不同的部分。 该方法还包括将至少一个部分表示为原语并存储原语的参数。 原语可以是多项式。