公开(公告)号：US08612368B2

公开(公告)日：2013-12-17

申请号：US13038086

申请日：2011-03-01

申请人： Douglas Ronald Burdick ,  Amol Ghoting ,  Rajasekar Krishnamurthy ,  Edwin Peter Dawson Pednault ,  Berthold Reinwald ,  Vikas Sindhwani ,  Shirish Tatikonda ,  Yuanyuan Tian ,  Shivakumar Vaithyanathan

发明人： Douglas Ronald Burdick ,  Amol Ghoting ,  Rajasekar Krishnamurthy ,  Edwin Peter Dawson Pednault ,  Berthold Reinwald ,  Vikas Sindhwani ,  Shirish Tatikonda ,  Yuanyuan Tian ,  Shivakumar Vaithyanathan

IPC分类号： G06F15/18

CPC分类号： G06F9/4843 ,  G06F8/427

摘要： Systems and methods for processing Machine Learning (ML) algorithms in a MapReduce environment are described. In one embodiment of a method, the method includes receiving a ML algorithm to be executed in the MapReduce environment. The method further includes parsing the ML algorithm into a plurality of statement blocks in a sequence, wherein each statement block comprises a plurality of basic operations (hops). The method also includes automatically determining an execution plan for each statement block, wherein at least one of the execution plans comprises one or more low-level operations (lops). The method further includes implementing the execution plans in the sequence of the plurality of the statement blocks.

摘要翻译： 描述了MapReduce环境中处理机器学习（ML）算法的系统和方法。 在一种方法的一个实施例中，该方法包括接收要在MapReduce环境中执行的ML算法。 该方法还包括将ML算法解析为序列中的多个语句块，其中每个语句块包括多个基本操作（跳）。 该方法还包括自动确定每个语句块的执行计划，其中至少一个执行计划包括一个或多个低级操作（lops）。 该方法还包括以多个语句块的顺序来实现执行计划。

公开(公告)号：US20120226639A1

公开(公告)日：2012-09-06

申请号：US13038086

申请日：2011-03-01

申请人： Douglas Ronald Burdick ,  Amol Ghoting ,  Rajasekar Krishnamurthy ,  Edwin Peter Dawson Pednault ,  Berthold Reinwald ,  Vikas Sindhwani ,  Shirish Tatikonda ,  Yuanyuan Tian ,  Shivakumar Vaithyanathan

发明人： Douglas Ronald Burdick ,  Amol Ghoting ,  Rajasekar Krishnamurthy ,  Edwin Peter Dawson Pednault ,  Berthold Reinwald ,  Vikas Sindhwani ,  Shirish Tatikonda ,  Yuanyuan Tian ,  Shivakumar Vaithyanathan

IPC分类号： G06F15/18

CPC分类号： G06F9/4843 ,  G06F8/427

摘要： Systems and methods for processing Machine Learning (ML) algorithms in a MapReduce environment are described. In one embodiment of a method, the method includes receiving a ML algorithm to be executed in the MapReduce environment. The method further includes parsing the ML algorithm into a plurality of statement blocks in a sequence, wherein each statement block comprises a plurality of basic operations (hops). The method also includes automatically determining an execution plan for each statement block, wherein at least one of the execution plans comprises one or more low-level operations (lops). The method further includes implementing the execution plans in the sequence of the plurality of the statement blocks.

摘要翻译： 描述了MapReduce环境中处理机器学习（ML）算法的系统和方法。 在一种方法的一个实施例中，该方法包括接收要在MapReduce环境中执行的ML算法。 该方法还包括将ML算法解析为序列中的多个语句块，其中每个语句块包括多个基本操作（跳）。 该方法还包括自动确定每个语句块的执行计划，其中至少一个执行计划包括一个或多个低级操作（lops）。 该方法还包括以多个语句块的顺序来实现执行计划。

公开(公告)号：US08359226B2

公开(公告)日：2013-01-22

申请号：US11335684

申请日：2006-01-20

申请人： Naoki Abe ,  Edwin Peter Dawson Pednault

发明人： Naoki Abe ,  Edwin Peter Dawson Pednault

IPC分类号： G06Q10/00

CPC分类号： G06Q30/0244 ,  G06Q10/06315 ,  G06Q10/0637 ,  G06Q30/02 ,  G06Q30/0201

摘要： A method of marketing optimization with respect to brand lifetime management formulates a problem of brand equity maximization utilizing Markov Decision Process (MDP) thereby casting brand equity management as a long term regard optimization problem in MDP, The marketing mix is optimized by formulating the mix as actions in MDP and, utilizing historical marketing and transaction data, aspects of the MDP are estimated.

公开(公告)号：US07519553B2

公开(公告)日：2009-04-14

申请号：US11743293

申请日：2007-05-02

申请人： Naoki Abe ,  James J. Bennett ,  David L. Jensen ,  Richard D. Lawrence ,  Prem Melville ,  Edwin Peter Dawson Pednault ,  Cezar Pendus ,  Chandan Karrem Reddy ,  Vincent Philip Thomas

发明人： Naoki Abe ,  James J. Bennett ,  David L. Jensen ,  Richard D. Lawrence ,  Prem Melville ,  Edwin Peter Dawson Pednault ,  Cezar Pendus ,  Chandan Karrem Reddy ,  Vincent Philip Thomas

IPC分类号： G06Q40/00

CPC分类号： G06Q40/00 ,  G06Q40/025 ,  G06Q40/12

摘要： The present invention employs data processing systems to handle debt collection by formulation the collections process as a Markov Decision Process with constrained resources, thus making it possible automatically to generate an optimal collections policy with respect to maximizing long-term expected return throughout the course of a collections process, subject to constraints on the available resources possibly in multiple organizations. This is accomplished by coupling data modeling and resource optimization within the constrained Markov Decision Process formulation and generating optimized rules based on constrained reinforcement learning process comprising applied on the basis of past historical data.

摘要翻译： 本发明采用数据处理系统来处理债务收集，通过将收集过程作为具有约束资源的马尔可夫决策过程来制定收款过程，从而使得可以自动生成关于在整个过程中最大化长期预期回报的最佳收款策略 收集过程，受到可能在多个组织中可用资源的限制。 这通过在约束马尔可夫决策过程公式中耦合数据建模和资源优化来实现，并且基于约束强化学习过程生成优化的规则，其包括基于过去的历史数据应用。

公开(公告)号：US20090030864A1

公开(公告)日：2009-01-29

申请号：US12204670

申请日：2008-09-04

申请人： Edwin Peter Dawson Pednault ,  Ramesh Natarajan

发明人： Edwin Peter Dawson Pednault ,  Ramesh Natarajan

IPC分类号： G06N5/02

CPC分类号： G06F17/30539 ,  G06K9/6219 ,  G06K9/6256 ,  G06K9/6282 ,  Y10S707/99942

摘要： The present invention generally relates to computer databases and, more particularly, to data mining and knowledge discovery. The invention specifically relates to a method for constructing segmentation-based predictive models, such as decision-tree classifiers, wherein data records are partitioned into a plurality of segments and separate predictive models are constructed for each segment. The present invention contemplates a computerized method for automatically building segmentation-based predictive models that substantially improves upon the modeling capabilities of decision trees and related technologies, and that automatically produces models that are competitive with, if not better than, those produced by data analysts and applied statisticians using traditional, labor-intensive statistical techniques. The invention achieves these properties by performing segmentation and multivariate statistical modeling within each segment simultaneously. Segments are constructed so as to maximize the accuracies of the predictive models within each segment. Simultaneously, the multivariate statistical models within each segment are refined so as to maximize their respective predictive accuracies.

摘要翻译： 本发明一般涉及计算机数据库，更具体地涉及数据挖掘和知识发现。 本发明具体涉及一种用于构建基于分段的预测模型的方法，例如决策树分类器，其中数据记录被划分成多个段，并且为每个段构建单独的预测模型。 本发明考虑了一种用于自动建立基于分段的预测模型的计算机化方法，其基本上改进了决策树和相关技术的建模能力，并且自动生成与数据分析者和 应用统计学家使用传统的劳动密集型统计技术。 本发明通过在每个段内同时执行分段和多变量统计建模来实现这些特性。 构建分段以最大化每个分段内的预测模型的准确性。 同时，对各段内的多元统计模型进行细化，使其各自的预测精度达到最大化。

公开(公告)号：US07072841B1

公开(公告)日：2006-07-04

申请号：US09302154

申请日：1999-04-29

申请人： Edwin Peter Dawson Pednault

发明人： Edwin Peter Dawson Pednault

IPC分类号： G06F17/60

CPC分类号： G06Q40/00 ,  G06F19/00 ,  G06Q40/08 ,  G06Q50/24

摘要： The invention considers a widely applicable method of constructing segmentation-based predictive models from data that permits constraints to be placed on the statistical estimation errors that can be tolerated with respect to various aspects of the models that are constructed. The present invention uses these statistical constraints in a closed-loop fashion to guide the construction of potential segments so as to produce segments that satisfy the statistical constraints whenever it is feasible to do so. The method is closed-loop in a sense that the statistical constraints are used in a manner that is analogous to an error signal in a feed-back control system, wherein the error signal is used to regulate the inputs to the process that is being controlled.

摘要翻译： 本发明考虑了一种广泛应用的从数据构建基于分段的预测模型的方法，所述数据允许将约束放置在关于构建的模型的各个方面可以容忍的统计估计误差上。 本发明以闭环方式使用这些统计约束来指导潜在段的构建，以便在可行时产生满足统计约束的段。 该方法是闭环的，其中统计约束以类似于反馈控制系统中的误差信号的方式使用，其中误差信号用于调节正被控制的过程的输入 。

公开(公告)号：US5970464A

公开(公告)日：1999-10-19

申请号：US926804

申请日：1997-09-10

申请人： Chidanand Vinayak Apte ,  Edna Grossman ,  Edwin Peter Dawson Pednault ,  Barry Kumin Rosen ,  Fateh Ali Tipu ,  Hsueh-ju Wang ,  Brian Frederick White

发明人： Chidanand Vinayak Apte ,  Edna Grossman ,  Edwin Peter Dawson Pednault ,  Barry Kumin Rosen ,  Fateh Ali Tipu ,  Hsueh-ju Wang ,  Brian Frederick White

IPC分类号： G06Q10/06 ,  G06F17/60

CPC分类号： G06Q40/08 ,  G06Q10/0635

摘要： A computer implemented method of underwriting profitability analysis delivers the analytic process to a wide cross section of insurance decision makers. The underwriting profitability analysis system leverages an existing investment in databases and improves underwriting business processes. Data mining techniques are applied to historical policy and claims to extract rules that describe policy holders with homogeneous claim frequency and severity characteristics. These rule sets are used to classify policy holders into distinct risk groups, each with its own set of characteristics, including pure premium. Breaking up a book of business into segments allows identification of sub-populations of policy holders that distinctly deviate from the expected normal pure premium. This identification allow the insurance business analysts to interactively adjust eligibility criteria and examine altered characteristics of the covered segments until satisfactory. The system is implemented on a client server using network centric language technology.

摘要翻译： 计算机实施的盈利能力分析方法将分析过程提供给广泛的保险决策者。 承销盈利能力分析系统利用现有的数据库投资，改善承销业务流程。 数据挖掘技术适用于历史政策和索赔，以提取描述具有均匀索赔频率和严重性特征的策略持有人的规则。 这些规则集用于将保单持有人分类为不同的风险群体，每个风险群体都有自己的一套特征，包括纯保费。 将业务分成几部分，可以确定明显偏离预期的正常纯价格的保单持有人的子种群。 这种识别允许保险业务分析人员交互地调整资格标准，并检查被覆盖部分的变更特征，直到令人满意。 该系统在使用网络中心语言技术的客户端服务器上实现。

公开(公告)号：US20090112927A1

公开(公告)日：2009-04-30

申请号：US11924840

申请日：2007-10-26

申请人： Upendra Chitnis ,  Christoph Lingenfelder ,  Edwin Peter Dawson Pednault

发明人： Upendra Chitnis ,  Christoph Lingenfelder ,  Edwin Peter Dawson Pednault

IPC分类号： G06F17/30

CPC分类号： G06Q10/087 ,  G06Q30/02

摘要： A process of transforming data residing in databases, such as relational databases, into forms suitable as input to data analysis tools, such as predictive modeling tools includes the steps of defining a business process problem to be solved and identifying data requirements. For example, the business process problem may relate to predicting a customer's propensity to make purchases in the future or a store's requirements for inventory in the future. In the process, a computer implemented method is used for automatically transforming data for data analysis such as predictive modeling. Database metadata that describe database tables, their interrelationships, dimensional information, fact tables and measures are accessed. A mining transformation profile is created to encapsulate aggregations and transformation on data stored in relational databases in order to convert the data to forms suitable for predictive mining tools. The mining transformation profile specifies data transformations relative to the data base metadata. Executable data transformation codes is then generated from the database metadata and the mining transformation profile. Execution of this code results in aggregation and transformation of data residing in a database for input to a data analysis tool such as a predictive modeling tool. The data transformation code can be used by, for example, the predictive modeling tool to generate an output that provides a solution to a business process problem.

摘要翻译： 将数据库（例如关系数据库）中驻留的数据转换为适合作为数据分析工具（例如预测建模工具）的输入的形式的过程包括以下步骤：定义要解决的业务流程问题并识别数据需求。 例如，业务流程问题可能与预测客户未来进行购买的倾向或商店对库存的需求有关。 在此过程中，使用计算机实现的方法来自动转换数据进行数据分析，如预测建模。 访问描述数据库表，它们的相互关系，维度信息，事实表和度量的数据库元数据。 创建挖掘转换配置文件以将聚合和变换封装在关系数据库中存储的数据上，以将数据转换为适合预测挖掘工具的表单。 挖掘转换配置文件指定相对于数据库元数据的数据转换。 然后从数据库元数据和挖掘转换配置文件生成可执行的数据转换代码。 执行此代码导致驻留在数据库中的数据的聚合和变换，以输入到诸如预测建模工具的数据分析工具。 数据转换代码可以由例如预测建模工具用于生成提供业务流程问题解决方案的输出。

公开(公告)号：US20080275800A1

公开(公告)日：2008-11-06

申请号：US11743293

申请日：2007-05-02

申请人： Naoki Abe ,  James J. Bennett ,  David L. Jensen ,  Richard D. Lawrence ,  Prem Melville ,  Edwin Peter Dawson Pednault ,  Cezar Pendus ,  Chandan Karrem Reddy ,  Vincent Philip Thomas

发明人： Naoki Abe ,  James J. Bennett ,  David L. Jensen ,  Richard D. Lawrence ,  Prem Melville ,  Edwin Peter Dawson Pednault ,  Cezar Pendus ,  Chandan Karrem Reddy ,  Vincent Philip Thomas

IPC分类号： G06Q40/00

CPC分类号： G06Q40/00 ,  G06Q40/025 ,  G06Q40/12

摘要： The present invention employs data processing systems to handle debt collection by formulation the collections process as a Markov Decision Process with constrained resources, thus making it possible automatically to generate an optimal collections policy with respect to maximizing long-term expected return throughout the course of a collections process, subject to constraints on the available resources possibly in multiple organizations. This is accomplished by coupling data modeling and resource optimization within the constrained Markov Decision Process formulation and generating optimized rules based on constrained reinforcement learning process comprising applied on the basis of past historical data.

摘要翻译： 本发明采用数据处理系统来处理债务收集，通过将收集过程作为具有约束资源的马尔可夫决策过程来制定收款过程，从而使得可以自动生成关于在整个过程中最大化长期预期回报的最佳收款策略 收集过程，受到可能在多个组织中可用资源的限制。 这通过在约束马尔可夫决策过程公式中耦合数据建模和资源优化来实现，并且基于约束强化学习过程生成优化的规则，其包括基于过去的历史数据应用。

公开(公告)号：US07246083B2

公开(公告)日：2007-07-17

申请号：US10401935

申请日：2003-03-27

申请人： Eric Bibelnieks ,  Mark Bullock ,  Michael Haydock ,  Mindy Schiller ,  Wayne Kugel ,  Edwin Peter Dawson Pednault ,  Nancy Soderquist ,  Harlan Crowder

发明人： Eric Bibelnieks ,  Mark Bullock ,  Michael Haydock ,  Mindy Schiller ,  Wayne Kugel ,  Edwin Peter Dawson Pednault ,  Nancy Soderquist ,  Harlan Crowder

IPC分类号： G06Q30/00

CPC分类号： G06Q30/02 ,  G06Q30/0244 ,  G06Q30/0245 ,  G06Q30/0247 ,  G06Q30/0249 ,  G06Q30/0254 ,  G06Q30/0277

摘要： A method, and system for targeted marketing of goods or services to specific customers is disclosed. Customers are analyzed based upon historical criteria; a promotional plan (a group of promotion events implemented or to be implemented over a particular time period) is analyzed to determine the effect of each promotion event on the other promotion events in the promotional plan; and, based on this analysis, the optimal promotion stream (a specific subset of the promotional plan to be sent to customers or a group of similar customers) is determined so as to maximize the ROI of the promotional plan as a whole.

摘要翻译： 披露了向特定客户进行商品或服务定向营销的方法和系统。 客户根据历史标准进行分析; 分析促销计划（在特定时间段内实施或将要实施的一组促销活动），以确定每个促销活动对促销计划中其他促销活动的影响; 并根据这一分析确定最佳促销流程（要发送给客户或一组类似客户的促销计划的特定子集），以便最大限度地提高整体促销计划的投资回报。