专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明授权

US09292798B2 Iterative active feature extraction 有权
标题翻译：迭代主动特征提取
公开(公告)号：US09292798B2
公开(公告)日：2016-03-22
申请号：US13723699
申请日：2012-12-21
申请人： International Business Machines Corporation
发明人： Christoph Lingenfelder , Pascal Pompey , Olivier Verscheure , Michael Wurst
IPC分类号： G06N5/02 , G06N99/00 , G06N5/04
CPC分类号： G06N99/005 , G06N5/02 , G06N5/025 , G06N5/043
摘要： Techniques for iterative feature extraction using domain knowledge are provided. In one aspect, a method for feature extraction is provided. The method includes the following steps. At least one query to predict at least one future value of a given value series based on a statistical model is received. At least two predictions of the future value are produced fulfilling at least the properties of 1) each being as probable as possible given the statistical model and 2) being mutually divert (in terms of numerical distance measure). A user is queried to select one of the predictions. The user may be queried for textual annotations for the predictions. The annotations may be used to identify additional covariates to create an extended set of covariates. The extended set of covariates may be used to improve the accuracy of the statistical model.
摘要翻译：提供了使用域知识进行迭代特征提取的技术。在一方面，提供了一种用于特征提取的方法。该方法包括以下步骤。接收至少一个基于统计模型来预测给定值序列的至少一个未来值的查询。至少产生两个未来价值的预测，至少满足1）的性质，每一个在统计模型中可能是可能的，2）相互转移（在数值距离测量方面）。查询用户以选择其中一个预测。可以查询用户的预测文本注释。注释可用于识别额外的协变量以创建扩展的一组协变量。扩展的协变量组可以用于提高统计模型的准确性。

2. 发明授权

US09235633B2 Processing data in a data warehouse 有权
标题翻译：处理数据仓库中的数据
公开(公告)号：US09235633B2
公开(公告)日：2016-01-12
申请号：US13705901
申请日：2012-12-05
申请人： International Business Machines Corporation
发明人： Iliyana P. Ivanova , Christoph Lingenfelder , Christoph H. Sieb , Simone Zerfass
IPC分类号： G06F17/30
CPC分类号： G06F17/30592 , G06F17/30333 , G06F17/30398 , G06F17/30563 , G06F17/30598 , G06F17/30867
摘要： Data of a database environment, which includes hierarchy information and a matrix of values, is processed. The hierarchy information includes at least two sets of identification codes and defines at least two groups of identification codes. The matrix of values includes at least two columns of identification values. At least one simple filter object is generated based on a user input. Each simple filter object defines an ad hoc group of identification codes selected from a respective one of the sets of identification codes. A filtered operation object that specifies an operation and at least one of the simple filter objects is generated based on a user input. Each of the ad hoc groups differs from each of the groups defined by the hierarchy information.
摘要翻译：处理数据库环境的数据，其中包括层次信息和值矩阵。层级信息包括至少两组识别码，并且定义至少两组识别码。值矩阵包括至少两列识别值。基于用户输入生成至少一个简单的过滤器对象。每个简单的过滤器对象定义了从识别代码集合中的相应一个中选择的一组自适应识别码。基于用户输入生成指定操作的过滤操作对象和至少一个简单过滤器对象。每个特设小组与由层级信息定义的每个组不同。

3. 发明授权

US09390377B2 Iterative active feature extraction 有权
公开(公告)号：US09390377B2
公开(公告)日：2016-07-12
申请号：US13785132
申请日：2013-03-05
申请人： International Business Machines Corporation
发明人： Christoph Lingenfelder , Pascal Pompey , Olivier Verscheure , Michael Wurst
IPC分类号： G06N5/02 , G06N99/00 , G06N5/04
CPC分类号： G06N99/005 , G06N5/02 , G06N5/025 , G06N5/043
摘要： Techniques for iterative feature extraction using domain knowledge are provided. In one aspect, a method for feature extraction is provided. The method includes the following steps. At least one query to predict at least one future value of a given value series based on a statistical model is received. At least two predictions of the future value are produced fulfilling at least the properties of 1) each being as probable as possible given the statistical model and 2) being mutually divert (in terms of numerical distance measure). A user is queried to select one of the predictions. The user may be queried for textual annotations for the predictions. The annotations may be used to identify additional covariates to create an extended set of covariates. The extended set of covariates may be used to improve the accuracy of the statistical model.

4. 发明申请

US20140258311A1 INSIGHT DETERMINATION AND EXPLANATION IN MULTI-DIMENSIONAL DATA SETS 有权
标题翻译：多维数据集中的明确确定和解释
公开(公告)号：US20140258311A1
公开(公告)日：2014-09-11
申请号：US13788681
申请日：2013-03-07
申请人： INTERNATIONAL BUSINESS MACHINES CORPORATION
发明人： Felix Hamborg , Alexander Lang , Christoph Lingenfelder
IPC分类号： G06F17/30
CPC分类号： G06F17/3053 , G06F17/30592
摘要： Techniques are disclosed for determining reasons underlying insights gleaned from multi-dimensional data. In one embodiment, a contingency table is accessed that represents multiple dimensions of the data, in order to identify one or more insights. One or more dimensions, other than the represented dimensions, are evaluated to identify one or more reasons underlying a first insight of the one or more insights, and the one or more reasons are output.
摘要翻译：披露技术来确定从多维数据中收集的见解的原因。在一个实施例中，访问表示数据的多个维度的意外事件表，以便识别一个或多个见解。评估一个或多个维度，而不是所表示的维度，以识别一个或多个见解的第一个洞察的一个或多个原因，并且输出一个或多个原因。

5. 发明申请

US20160034552A1 PROCESSING DATA IN A DATA WAREHOUSE 有权
标题翻译：在数据仓库中处理数据
公开(公告)号：US20160034552A1
公开(公告)日：2016-02-04
申请号：US14879222
申请日：2015-10-09
申请人： International Business Machines Corporation
发明人： Iliyana P. Ivanova , Christoph Lingenfelder , Christoph H. Sieb , Simone Zerfass
IPC分类号： G06F17/30
CPC分类号： G06F17/30592 , G06F17/30333 , G06F17/30398 , G06F17/30563 , G06F17/30598 , G06F17/30867
摘要： Processing data of a data warehouse is provided and includes receiving, by a processing device, user input to create simple filter objects. Each filter object defines an ad hoc subset of a respective dimension of a dimension table of the data warehouse. User input is received to create a filtered operation object that specifies an operation and a plurality of the simple filter objects. The ad hoc subset differs from all subsets defined in the dimension table.
摘要翻译：提供数据仓库的处理数据，并且包括由处理设备接收用户输入以创建简单的过滤器对象。每个过滤器对象定义数据仓库的维度表的相应维度的ad hoc子集。接收用户输入以创建指定操作的过滤的操作对象和多个简单的过滤对象。 ad hoc子集与维度表中定义的所有子集不同。

6. 发明授权

US09218396B2 Insight determination and explanation in multi-dimensional data sets 有权
标题翻译：多维数据集中的洞察力确定与解释
公开(公告)号：US09218396B2
公开(公告)日：2015-12-22
申请号：US14224334
申请日：2014-03-25
申请人： International Business Machines Corporation
发明人： Felix Hamborg , Alexander Lang , Christoph Lingenfelder
IPC分类号： G06F17/30
CPC分类号： G06F17/3053 , G06F17/30592
摘要： Techniques are disclosed for determining reasons underlying insights gleaned from multi-dimensional data. In one embodiment, a contingency table is accessed that represents multiple dimensions of the data, in order to identify one or more insights. One or more dimensions, other than the represented dimensions, are evaluated to identify one or more reasons underlying a first insight of the one or more insights, and the one or more reasons are output.
摘要翻译：披露技术来确定从多维数据中收集的见解的原因。在一个实施例中，访问表示数据的多个维度的意外事件表，以便识别一个或多个见解。评估一个或多个维度，而不是所表示的维度，以识别一个或多个见解的第一个洞察的一个或多个原因，并且输出一个或多个原因。

7. 发明申请

US20140180992A1 Iterative Active Feature Extraction 有权
标题翻译：迭代主动特征提取
公开(公告)号：US20140180992A1
公开(公告)日：2014-06-26
申请号：US13785132
申请日：2013-03-05
申请人： INTERNATIONAL BUSINESS MACHINES CORPORATION
发明人： Christoph Lingenfelder , Pascal Pompey , Olivier Verscheure , Michael Wurst
IPC分类号： G06N5/02
CPC分类号： G06N99/005 , G06N5/02 , G06N5/025 , G06N5/043
摘要： Techniques for iterative feature extraction using domain knowledge are provided. In one aspect, a method for feature extraction is provided. The method includes the following steps. At least one query to predict at least one future value of a given value series based on a statistical model is received. At least two predictions of the future value are produced fulfilling at least the properties of 1) each being as probable as possible given the statistical model and 2) being mutually divert (in terms of numerical distance measure). A user is queried to select one of the predictions. The user may be queried for textual annotations for the predictions. The annotations may be used to identify additional covariates to create an extended set of covariates. The extended set of covariates may be used to improve the accuracy of the statistical model.
摘要翻译：提供了使用域知识进行迭代特征提取的技术。在一方面，提供了一种用于特征提取的方法。该方法包括以下步骤。接收至少一个基于统计模型来预测给定值序列的至少一个未来值的查询。至少产生两个未来价值的预测，至少满足1）的性质，每一个在统计模型中可能是可能的，2）相互转移（在数值距离测量方面）。查询用户以选择其中一个预测。可以查询用户的预测文本注释。注释可用于识别额外的协变量以创建扩展的一组协变量。扩展的协变量组可以用于提高统计模型的准确性。

8. 发明申请

US20130144833A1 PROCESSING DATA IN A DATA WAREHOUSE 有权
标题翻译：在数据仓库中处理数据
公开(公告)号：US20130144833A1
公开(公告)日：2013-06-06
申请号：US13705901
申请日：2012-12-05
申请人： International Business Machines Corporation
发明人： Iliyana P. Ivanova , Christoph Lingenfelder , Christoph H. Sieb , Simone Zerfass
IPC分类号： G06F17/30
CPC分类号： G06F17/30592 , G06F17/30333 , G06F17/30398 , G06F17/30563 , G06F17/30598 , G06F17/30867
摘要： Data of a database environment, which includes hierarchy information and a matrix of values, is processed. The hierarchy information includes at least two sets of identification codes and defines at least two groups of identification codes. The matrix of values includes at least two columns of identification values. At least one simple filter object is generated based on a user input. Each simple filter object defines an ad hoc group of identification codes selected from a respective one of the sets of identification codes. A filtered operation object that specifies an operation and at least one of the simple filter objects is generated based on a user input. Each of the ad hoc groups differs from each of the groups defined by the hierarchy information.
摘要翻译：处理数据库环境的数据，其中包括层次信息和值矩阵。层级信息包括至少两组识别码，并且定义至少两组识别码。值矩阵包括至少两列识别值。基于用户输入生成至少一个简单的过滤器对象。每个简单的过滤器对象定义了从识别代码集合中的相应一个中选择的一组自适应识别码。基于用户输入生成指定操作的过滤操作对象和至少一个简单过滤器对象。每个特设小组与由层级信息定义的每个组不同。

9. 发明授权

US09639596B2 Processing data in a data warehouse 有权
公开(公告)号：US09639596B2
公开(公告)日：2017-05-02
申请号：US14879222
申请日：2015-10-09
申请人： International Business Machines Corporation
发明人： Iliyana P. Ivanova , Christoph Lingenfelder , Christoph H. Sieb , Simone Zerfass
IPC分类号： G06F17/30
CPC分类号： G06F17/30592 , G06F17/30333 , G06F17/30398 , G06F17/30563 , G06F17/30598 , G06F17/30867
摘要： Processing data of a data warehouse is provided and includes receiving, by a processing device, user input to create simple filter objects. Each filter object defines an ad hoc subset of a respective dimension of a dimension table of the data warehouse. User input is received to create a filtered operation object that specifies an operation and a plurality of the simple filter objects. The ad hoc subset differs from all subsets defined in the dimension table.

10. 发明授权

US09223832B2 Insight determination and explanation in multi-dimensional data sets 有权
公开(公告)号：US09223832B2
公开(公告)日：2015-12-29
申请号：US13788681
申请日：2013-03-07
申请人： International Business Machines Corporation
发明人： Felix Hamborg , Alexander Lang , Christoph Lingenfelder
IPC分类号： G06F17/30
CPC分类号： G06F17/3053 , G06F17/30592
摘要： Techniques are disclosed for determining reasons underlying insights gleaned from multi-dimensional data. In one embodiment, a contingency table is accessed that represents multiple dimensions of the data, in order to identify one or more insights. One or more dimensions, other than the represented dimensions, are evaluated to identify one or more reasons underlying a first insight of the one or more insights, and the one or more reasons are output.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式