METHOD OF TRAINING A NEURAL NETWORK AND A NEURAL NETWORK TRAINED ACCORDING TO THE METHOD
    1.
    发明申请
    METHOD OF TRAINING A NEURAL NETWORK AND A NEURAL NETWORK TRAINED ACCORDING TO THE METHOD 审中-公开
    训练神经网络的方法和根据该方法训练的神经网络

    公开(公告)号:US20080301075A1

    公开(公告)日:2008-12-04

    申请号:US11936756

    申请日:2007-11-07

    IPC分类号: G06N3/08

    CPC分类号: G06N3/08

    摘要: A neural network comprises trained interconnected neurons. The neural network is configured to constrain the relationship between one or more inputs and one or more outputs of the neural network so the relationships between them are consistent with expectations of the relationships; and/or the neural network is trained by creating a set of data comprising input data and associated outputs that represent archetypal results and providing real exemplary input data and associated output data and the created data to neural network. The real exemplary output data and the created associated output data is compared to the actual output of the neural network, which is adjusted to create a best fit to the real exemplary data and the created data.

    摘要翻译: 神经网络包括训练有素的相互关联的神经元。 神经网络被配置为约束神经网络的一个或多个输入和一个或多个输出之间的关系,使得它们之间的关系与关系的期望一致; 和/或通过创建一组数据来训练神经网络,该组数据包括表示原型结果的输入数据和相关联的输出,并且向神经网络提供真正的示例性输入数据和相关联的输出数据以及所创建的数据。 将真正的示例性输出数据和所创建的相关联的输出数据与神经网络的实际输出进行比较,神经网络被调整以创建与实际示例性数据和创建的数据的最佳拟合。

    Method of training a neural network and a neural network trained according to the method
    2.
    发明申请
    Method of training a neural network and a neural network trained according to the method 审中-公开
    根据该方法训练神经网络和神经网络的方法

    公开(公告)号:US20050149463A1

    公开(公告)日:2005-07-07

    申请号:US10976167

    申请日:2004-10-28

    IPC分类号: G06F15/18 G06G7/00 G06N3/08

    CPC分类号: G06N3/08

    摘要: A neural network comprises trained interconnected neurons. The neural network is configured to constrain the relationship between one or more inputs and one or more outputs of the neural network so the relationships between them are consistent with expectations of the relationships; and/or the neural network is trained by creating a set of data comprising input data and associated outputs that represent archetypal results and providing real exemplary input data and associated output data and the created data to neural network. The real exemplary output data and the created associated output data is compared to the actual output of the neural network, which is adjusted to create a best fit to the real exemplary data and the created data.

    摘要翻译: 神经网络包括训练有素的相互关联的神经元。 神经网络被配置为约束神经网络的一个或多个输入和一个或多个输出之间的关系,使得它们之间的关系与关系的期望一致; 和/或通过创建一组数据来训练神经网络,该组数据包括表示原型结果的输入数据和相关联的输出,并且向神经网络提供真正的示例性输入数据和相关联的输出数据以及所创建的数据。 将真正的示例性输出数据和所创建的相关联的输出数据与神经网络的实际输出进行比较,神经网络被调整以创建与实际示例性数据和创建的数据的最佳拟合。

    System and method for identifying extreme behavior in elements of a network
    3.
    发明申请
    System and method for identifying extreme behavior in elements of a network 有权
    用于识别网络元素中的极端行为的系统和方法

    公开(公告)号:US20050138463A1

    公开(公告)日:2005-06-23

    申请号:US10965670

    申请日:2004-10-14

    摘要: A system for identifying extreme behavior in elements of a network comprises a profiler and a collator. The profiler and the collator perform a method of identifying extreme behavior in the network elements. The profiler maintains one or more group profiles of network elements. Each group profile is associated with a plurality of network elements. The profiler accumulates values of a first function of the contents of an input data stream over a first period of time for each group profile. The input data stream includes at least one field containing a network element reference. The accumulated values of each group profile are compared with a corresponding collation threshold. The collator creates a collation instance for each group profile that reaches the collation threshold. Each collation instance creates a plurality of collation profiles. Each collation profile is associated with one or more network elements from the plurality of network elements corresponding to the group profile that caused the creation of the collation instance. The collator instance accumulates values of a second function of the contents of the input data stream for each collation profile over a second period of time. Extreme behavior of network elements is identified from the accumulated values of the collation profiles.

    摘要翻译: 用于识别网络元件中的极端行为的系统包括分析器和整理器。 分析器和整理器执行一种识别网络元素中的极端行为的方法。 分析器维护网络元素的一个或多个组简档。 每个组简档与多个网络元件相关联。 分析器在每个组简档的第一时间段内积累输入数据流的内容的第一函数的值。 输入数据流包括至少一个包含网元参考的字段。 将每个组轮廓的累积值与对应的归类阈值进行比较。 整理器为达到排序规则阈值的每个组配置文件创建一个排序规则实例。 每个排序规则实例创建多个排序规则。 每个核对简档与来自对应于引起对照实例的创建的组简档的多个网络元件中的一个或多个网络元件相关联。 整理器实例在第二时间段内累积用于每个核对简档的输入数据流的内容的第二函数的值。 网络元素的极端行为是从排序规则的累积值中确定出来的。

    Automated performance monitoring and adaptation system
    4.
    发明申请
    Automated performance monitoring and adaptation system 审中-公开
    自动化性能监测与适应系统

    公开(公告)号:US20050154688A1

    公开(公告)日:2005-07-14

    申请号:US10987451

    申请日:2004-11-12

    IPC分类号: H04W24/00 H04W24/04 G06F15/18

    CPC分类号: H04W24/04

    摘要: An event detection system with an automatic performance monitoring and adaptation system is disclosed. The system includes an event detection engine and a performance assessor. The event detection engine generates an alert if the specified event is suspected. An alert investigation team investigates if the alert is real of false. The performance assessor is configured to monitor the rate at which alerts and/or false alerts are generated by the event detection engine and to perform certain actions if the rate of alerts and/or false alerts falls outside a configurable range or crosses a threshold.

    摘要翻译: 公开了一种具有自动性能监测和调整系统的事件检测系统。 该系统包括事件检测引擎和性能评估器。 如果涉嫌指定事件,事件检测引擎会生成警报。 一个警戒调查小组调查警报是否是真实的。 性能评估器被配置为监视事件检测引擎产生警报和/或假警报的速率,并且如果警报和/或虚假警报的速率超出可配置范围或跨越阈值,则执行某些动作。

    Method and system for detecting change in data streams
    5.
    发明授权
    Method and system for detecting change in data streams 有权
    检测数据流变化的方法和系统

    公开(公告)号:US07620533B2

    公开(公告)日:2009-11-17

    申请号:US11654800

    申请日:2007-01-18

    IPC分类号: G06F17/10

    CPC分类号: G06K9/6268 G05B23/0254

    摘要: A system for detecting change in a data stream comprising a distribution maintenance engine, a difference determining means and an alert generation engine is disclosed. The system detects change in the alert stream by the distribution maintenance engine maintaining a short term distribution that models the data stream and maintaining a long term distribution that models the data stream. The difference determining means determines the difference between the short term distribution and the long term distribution. The alert generation engine applies a statistical measure to the difference and generates an alert if the measure of the difference exceeds a threshold.

    摘要翻译: 公开了一种用于检测包括分发维护引擎,差分确定装置和警报生成引擎的数据流的变化的系统。 该系统检测分发维护引擎维护警报流的变化,保持对数据流进行建模的短期分布,并维护对数据流建模的长期分布。 差分确定装置确定短期分布与长期分布之间的差异。 警报生成引擎对差异应用统计度量,并且如果差异的度量超过阈值则产生警报。

    Configurable profiling of data
    6.
    发明授权
    Configurable profiling of data 有权
    可配置的数据分析

    公开(公告)号:US07471780B2

    公开(公告)日:2008-12-30

    申请号:US10509600

    申请日:2003-03-28

    摘要: In one embodiment, a method of configurably profiling data uses system comprising a pre-processor, a profiler, a post-processor and database. The pre-processor receives data and feedback data and performs configurable first calculations thereon to create data relating to profiling features. The profiler summarizes the profiling features data over a length of time by performing configurable second calculations thereon to create summarized data relating to profiled features. The post-processor performs configurable third calculations the profiled features data to create the feedback data and a profiled output data stream for further processing.

    摘要翻译: 在一个实施例中,可配置地分析数据的方法使用包括预处理器,分析器,后处理器和数据库的系统。 预处理器接收数据和反馈数据,并对其进行可配置的第一次计算,以创建与分析特征有关的数据。 分析器通过对其进行可配置的第二次计算来总结一段时间内的分析特征数据,以创建与分析特征相关的汇总数据。 后处理器执行可配置的第三次计算,分析特征数据以创建反馈数据和用于进一步处理的分析输出数据流。

    Configurable profiling of data
    7.
    发明申请
    Configurable profiling of data 有权
    可配置的数据分析

    公开(公告)号:US20050246182A1

    公开(公告)日:2005-11-03

    申请号:US10509600

    申请日:2003-03-28

    IPC分类号: H04M3/22 H04M15/00 G06F17/60

    摘要: In one embodiment, a method of configurably profiling data uses system comprising a pre-processor, a profiler, a post-processor and database. The pre-processor receives data and feedback data and performs configurable first calculations thereon to create data relating to profiling features. The profiler summarises the profiling features data over a length of time by performing configurable second calculations thereon to create summarised data relating to profiled features. The post-processor performs configurable third calculations the profiled features data to create the feedback data and a profiled output data stream for further processing.

    摘要翻译: 在一个实施例中,可配置地分析数据的方法使用包括预处理器,分析器,后处理器和数据库的系统。 预处理器接收数据和反馈数据,并对其进行可配置的第一次计算,以创建与分析特征有关的数据。 分析器通过对其进行可配置的第二次计算来总结一段时间内的分析特征数据,以创建与分析特征相关的汇总数据。 后处理器执行可配置的第三次计算,分析特征数据以创建反馈数据和用于进一步处理的分析输出数据流。

    Method and system for detecting change in data streams
    8.
    发明申请
    Method and system for detecting change in data streams 审中-公开
    检测数据流变化的方法和系统

    公开(公告)号:US20050149299A1

    公开(公告)日:2005-07-07

    申请号:US10971715

    申请日:2004-10-22

    IPC分类号: G05B23/02 G06F17/10

    CPC分类号: G06K9/6268 G05B23/0254

    摘要: A system for detecting change in a data stream comprising a distribution maintenance engine, a difference determining means and an alert generation engine is disclosed. The system detects change in the alert stream by the distribution maintenance engine maintaining a short term distribution that models the data stream and maintaining a long term distribution that models the data stream. The difference determining means determines the difference between the short term distribution and the long term distribution. The alert generation engine applies a statistical measure to the difference and generates an alert if the measure of the difference exceeds a threshold.

    摘要翻译: 公开了一种用于检测包括分发维护引擎,差分确定装置和警报生成引擎的数据流的变化的系统。 该系统检测分发维护引擎维护警报流的变化,保持对数据流进行建模的短期分布,并维护对数据流建模的长期分布。 差分确定装置确定短期分布与长期分布之间的差异。 警报生成引擎对差异应用统计度量,并且如果差异的度量超过阈值则产生警报。

    System and method for identifying extreme behavior in elements of a network
    9.
    发明授权
    System and method for identifying extreme behavior in elements of a network 有权
    用于识别网络元素中的极端行为的系统和方法

    公开(公告)号:US07631355B2

    公开(公告)日:2009-12-08

    申请号:US10965670

    申请日:2004-10-14

    IPC分类号: G06F11/00

    摘要: A system for identifying extreme behavior in elements of a network comprises a profiler and a collator. The profiler and the collator perform a method of identifying extreme behavior in the network elements. The profiler maintains one or more group profiles of network elements. Each group profile is associated with a plurality of network elements. The profiler accumulates values of a first function of the contents of an input data stream over a first period of time for each group profile. The input data stream includes at least one field containing a network element reference. The accumulated values of each group profile are compared with a corresponding collation threshold. The collator creates a collation instance for each group profile that reaches the collation threshold. Each collation instance creates a plurality of collation profiles. Each collation profile is associated with one or more network elements from the plurality of network elements corresponding to the group profile that caused the creation of the collation instance. The collator instance accumulates values of a second function of the contents of the input data stream for each collation profile over a second period of time. Extreme behavior of network elements is identified from the accumulated values of the collation profiles.

    摘要翻译: 用于识别网络元件中的极端行为的系统包括分析器和整理器。 分析器和整理器执行一种识别网络元素中的极端行为的方法。 分析器维护网络元素的一个或多个组简档。 每个组简档与多个网络元件相关联。 分析器在每个组简档的第一时间段内积累输入数据流的内容的第一函数的值。 输入数据流包括至少一个包含网元参考的字段。 将每个组轮廓的累积值与对应的归类阈值进行比较。 整理器为达到排序规则阈值的每个组配置文件创建一个排序规则实例。 每个排序规则实例创建多个排序规则。 每个核对简档与来自对应于引起对照实例的创建的组简档的多个网络元件中的一个或多个网络元件相关联。 整理器实例在第二时间段内累积用于每个核对简档的输入数据流的内容的第二函数的值。 网络元素的极端行为是从排序规则的累积值中确定出来的。

    HIERARCHICAL SYSTEM AND METHOD FOR ANALYZING DATA STREAMS
    10.
    发明申请
    HIERARCHICAL SYSTEM AND METHOD FOR ANALYZING DATA STREAMS 审中-公开
    分析数据流的分层系统和方法

    公开(公告)号:US20090164761A1

    公开(公告)日:2009-06-25

    申请号:US12340504

    申请日:2008-12-19

    IPC分类号: G06F9/30

    CPC分类号: H04M3/2281 H04L41/06

    摘要: A method for analyzing data streams comprises receiving a data stream, conducting a first analysis of the data stream for a possible target activity, and if a possible target activity is indicated generating a first alert. If the first alert is generated, a second analysis for the possible target activity is conducted to determine whether the target activity is indicated in the data stream with a high degree of certainty. If a possible target activity is indicated by the second analysis, a second alert is generated and provided to an external system for action.

    摘要翻译: 用于分析数据流的方法包括接收数据流,对可能的目标活动进行数据流的第一次分析,以及如果指示可能的目标活动产生第一警报。 如果产生了第一个警报,则进行可能的目标活动的第二次分析,以确定目标活动是否以高度确定性在数据流中指示。 如果通过第二次分析指示可能的目标活动,则生成第二个警报并提供给外部系统以进行操作。