公开(公告)号：US09798980B2

公开(公告)日：2017-10-24

申请号：US14627064

申请日：2015-02-20

Applicant: InSnap, Inc.

Inventor： Mohammad Sabah ,  Mohammad Iman Sadreddin ,  Shafaq Abdullah

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  G06N5/04 ,  G06F17/30 ,  G06N5/02 ,  G06Q30/02

CPC classification number: G06N5/048 ,  G06F17/30038 ,  G06F17/30247 ,  G06F17/30256 ,  G06F17/30277 ,  G06F17/3028 ,  G06F17/30377 ,  G06F17/3053 ,  G06F17/30598 ,  G06F17/30958 ,  G06F17/30991 ,  G06N5/02 ,  G06Q30/0269

Abstract: Techniques disclosed herein describe inferring user interests based on metadata of a plurality of multimedia objects captured by a plurality of users. An analysis tool receives, for each of the users, metadata describing each multimedia object in the plurality of objects associated with that user. Each multimedia object includes one or more attributes imputed to that object based on the metadata. The analysis tool identifies one or more concepts from the one or more attributes. Each concept includes at least a first attribute that co-occurs with a second attribute imputed to a first multimedia object. The analysis tool associates a first one of the plurality of users with at least one of the concepts based on the attributes imputed to multimedia objects associated with the first one of the plurality of users.

公开(公告)号：US09792561B2

公开(公告)日：2017-10-17

申请号：US14812650

申请日：2015-07-29

Applicant: FUJITSU LIMITED

Inventor： Makiko Konoshima ,  Yui Noma

IPC: G06N99/00 ,  G06F17/16 ,  G06N7/06

CPC classification number: G06N99/005 ,  G06F17/16 ,  G06N7/06

Abstract: A learning method includes: counting any one of or some of the number of labels added to each of feature amount vectors included in a learning data set, the number of types of the label, the number of feature amount vectors added with the same label, and the number of data pairs used for learning of a hyperplane, by a processor; first selecting, according to a result of the counting, one or more generation methods from a plurality of previously stored generation methods that generate the data pairs from the learning data set, by the processor; generating, using the selected generation methods, the data pairs from the feature amount vectors included in the learning data set, by the processor; and first learning, using the generated data pairs, the hyperplane that divides a feature amount vector space, by the processor.

公开(公告)号：US09672193B2

公开(公告)日：2017-06-06

申请号：US14217707

申请日：2014-03-18

Applicant: SAS Institute Inc.

Inventor： Christian Macaro ,  Jan Chvosta ,  Mark Roland Little

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  G06F17/18 ,  G06N7/00

CPC classification number: G06F17/18 ,  G06N7/005

Abstract: Various embodiments are directed to techniques for selecting a subset of a set of simulated samples. A computer-program product including instructions to cause a computing device to order a plurality of UPDFs by UPDF value, wherein the plurality of UPDFs is associated with a chain of draws of a set of simulated samples, wherein each draw comprises multiple parameters and the UPDF values map to parameter values of the parameters; select a subset of the plurality of UPDFs based on the subset of the plurality of UPDFs having UPDF values within a range corresponding to a range of parameter values to include in a subset of the set of simulated samples; and transmit an indication of a draw comprising parameters having parameter values to include in the subset of the set of simulated samples, wherein the indication identifies the draw by associated UPDF. Other embodiments are described and claimed.

公开(公告)号：US09576326B2

公开(公告)日：2017-02-21

申请号：US14298777

申请日：2014-06-06

Applicant: LinkedIn Corporation

Inventor： Christina Yip Chung ,  Allen Blue

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  G06Q50/00 ,  G06Q10/10

CPC classification number: G06N7/005 ,  G06N99/005 ,  G06Q10/101 ,  G06Q50/01 ,  H04L67/22

Abstract: Techniques for identification of a propagator-type leader in a social network are described. According to various embodiments, a specific content item posted by a particular actor of a plurality of actors and interactions by other actors of the plurality of actors with the specific content item are identified. A leadership score associated with the particular actor is calculated, the leadership score indicating a propensity of the particular actor to spread information among the plurality of actors of the online social network service. The particular actor is then classified as an information propagator among the plurality of actors of the online social network service, based on the calculated leadership score.

Abstract translation: 描述了用于在社交网络中识别传播者类型的领导者的技术。 根据各种实施例，识别由多个角色的特定演员发布的特定内容项目以及具有特定内容项目的多个演员的其他演员的交互。 计算与特定演员相关联的领导分数，领导分数表示特定演员在在线社交网络服务的多个角色之间传播信息的倾向。 然后根据所计算的领导分数将特定的演员分类为在线社交网络服务的多个角色中的信息传播者。

公开(公告)号：US09519867B1

公开(公告)日：2016-12-13

申请号：US13665099

申请日：2012-10-31

Applicant: Sprint Communications Company L.P.

Inventor： Raymond Emilio Reeves ,  Jason Norwood Ward ,  Tuan Quoc Tran ,  Lauriann Lowe Hebb

IPC: G06N5/02 ,  G06F17/00 ,  G06N7/06 ,  G06F17/30

CPC classification number: G06N7/06 ,  G06F17/30289 ,  G06F17/30867

Abstract: Systems, methods, and computer-readable media for optimizing a user experience are provided. The method includes optimizing a user experience using clusters, user preferences, or a combination thereof. Clusters may be created based on, for example, user behaviors, or actions, exhibited by a user. User preferences may be established for each cluster in order to further customize the clusters. The clusters may be continuously monitored such that if changes are necessary they may be immediately applied such as a user exhibited different behavior and requiring association with a new cluster. This information, or clustering, may be utilized to predict user satisfaction such that more positive user experiences are encountered and negative user experiences are, to the extent possible, avoided, or at least lessened.

Abstract translation: 提供了用于优化用户体验的系统，方法和计算机可读介质。 该方法包括使用群集优化用户体验，用户偏好或其组合。 可以基于例如用户展示的用户行为或动作来创建群集。 可以为每个集群建立用户偏好，以进一步定制集群。 可以连续地监视集群，使得如果需要进行改变，则可以立即应用它们，例如用户展示不同的行为并且要求与新的集群关联。 可以利用这种信息或聚类来预测用户的满意度，以便遇到更积极的用户体验，并且在可能的情况下，消极的用户体验被避免或至少减轻。

公开(公告)号：US09477966B2

公开(公告)日：2016-10-25

申请号：US13335611

申请日：2011-12-22

Applicant: Edward L. Hewett ,  Eric V. Anderson ,  William Brandon George ,  Kevin G. Smith

Inventor： Edward L. Hewett ,  Eric V. Anderson ,  William Brandon George ,  Kevin G. Smith

IPC: G06N7/02 ,  G06N7/06 ,  G06Q30/02 ,  G06N7/04 ,  G06N5/04 ,  G06N99/00 ,  G06N7/00

CPC classification number: G06Q30/0242 ,  G06N5/04 ,  G06N5/048 ,  G06N7/005 ,  G06N7/04 ,  G06N99/005

Abstract: A method, system, and computer-readable storage medium are disclosed for accurately estimating the audience of digital content. A plurality of user interactions occurring on a digital device are received, wherein digital content is displayed on the digital device. Based on the plurality of user interactions, a total number of unique viewers of the digital content on the digital device is estimated.

Abstract translation: 公开了一种用于准确估计数字内容的观众的方法，系统和计算机可读存储介质。 接收在数字设备上发生的多个用户交互，其中数字内容被显示在数字设备上。 基于多个用户交互，估计数字设备上的数字内容的唯一观看者的总数。

公开(公告)号：US09443198B1

公开(公告)日：2016-09-13

申请号：US14192009

申请日：2014-02-27

Applicant: Amazon Technologies, Inc.

Inventor： Shiv Naga Prasad Vitaladevuni

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  G06N5/04 ,  G06N7/00 ,  G06N99/00 ,  G06N5/02 ,  G06K9/62

CPC classification number: G06N5/048 ,  G06K9/6256 ,  G06K9/6257 ,  G06K9/6269 ,  G06K9/6281 ,  G06N5/025 ,  G06N7/005 ,  G06N99/005

Abstract: Features are disclosed for detecting an event in input data using a cascade-based detection system. Detection of the event may be triggered at any stage of the cascade, and subsequent stages of the cascade are not reached in such cases. Individual stages of the cascade may be associated with detection thresholds for use in triggering detection of the event. The sequence of stages may be selected based on some observed or desired operational characteristic, such as latency or operational cost. In addition, the cascade may be modified or updated based on data received from client devices. The data may relate to measurements and determinations made during real-world use of the cascade to detect events in input data.

Abstract translation: 公开了用于使用基于级联的检测系统来检测输入数据中的事件的特征。 可以在级联的任何阶段触发事件的检测，并且在这种情况下不能达到级联的后续阶段。 级联的各个阶段可以与用于触发事件检测的检测阈值相关联。 可以基于一些观察到的或期望的操作特性（例如延迟或操作成本）来选择阶段的顺序。 此外，可以基于从客户端设备接收的数据来修改或更新级联。 该数据可能涉及在实际使用级联期间进行的测量和确定以检测输入数据中的事件。

公开(公告)号：US09342790B1

公开(公告)日：2016-05-17

申请号：US14591403

申请日：2015-01-07

Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor： Aaron K. Baughman ,  Christian Eggenberger ,  Peter K. Malkin ,  Mauro Marzorati

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  G06N7/00

CPC classification number: G06N7/005 ,  G06N5/043

Abstract: A forecasting cohort includes a first set of forecasting algorithms and a second set of forecasting algorithms. An initial confidence level and a half-life of each of the first set of forecasting algorithms and the second set of forecasting algorithms are determined. A half-life weight for each of the first set of forecasting algorithms and the second set of forecasting algorithms at a subsequent time are determined, such that the half-life weights decrease an effect of a forecasting algorithm as time elapses. A combined confidence level of the forecasting cohort at the subsequent time is determined and used to adjust resource usage.

Abstract translation: 预测队列包括第一组预测算法和第二组预测算法。 确定第一组预测算法和第二组预测算法中的每一个的初始置信水平和半衰期。 确定第一组预测算法和随后时间的第二组预测算法的半衰期权重，使得半衰期权重随着时间流逝降低预测算法的影响。 确定随后时间的预测队列的组合置信水平，并用于调整资源使用。

公开(公告)号：US09224098B2

公开(公告)日：2015-12-29

申请号：US13999071

申请日：2014-01-10

Applicant: Optibrium Ltd

Inventor： Matthew D. Segall ,  Iskander Yusof

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  G06N7/00 ,  G06N5/04 ,  G06Q10/04

CPC classification number: G06N7/005 ,  G06N5/04 ,  G06Q10/04

Abstract: Methods, software, products and systems used to support decision making in complex multidimensional problem environments. Methods, software, products and systems to prioritize solutions for selection based upon selection criteria and available data regarding the possible solutions. The methods achieve a robust approach to determine the sensitivity of a selection to a multi-parameter profile of selection criteria and the importance of such criteria.

Abstract translation: 用于支持复杂多维问题环境中决策的方法，软件，产品和系统。 方法，软件，产品和系统，以根据选择标准和可能的解决方案的可用数据优先选择解决方案。 这些方法实现了一种稳健的方法来确定选择对选择标准的多参数分布的灵敏度以及这些标准的重要性。

公开(公告)号：US09191404B2

公开(公告)日：2015-11-17

申请号：US13910182

申请日：2013-06-05

Applicant: Cisco Technology, Inc.

Inventor： Clinton Grant

IPC: G06F9/44 ,  G06N7/02 ,  G06N7/06 ,  H04L29/06 ,  H04L12/721 ,  H04L12/859

CPC classification number: H04L63/1441 ,  H04L45/38 ,  H04L47/2475

Abstract: Presented herein are probabilistic flow management techniques in which flow objects are probabilistically evaluated in view of the current contents of a flow table to determine if the flow object should be added to the flow table. An untrusted packet flow may be received at a feature or function of a networking device. The feature initiates addition of an untrusted flow object corresponding to the untrusted packet flow into a flow table. A probabilistic flow management mechanism determines if the number of untrusted flow objects in the flow table is below a predetermined lower limit. If the number of untrusted flow objects exceeds the lower limit and prior to addition of the untrusted flow object into the flow table, the probabilistic flow management mechanism probabilistically determines if the untrusted flow object may be added to the flow table.

Abstract translation: 这里呈现的是概率流管理技术，其中根据流表的当前内容来概率地评估流对象，以确定流对象是否应该添加到流表中。 可以在网络设备的特征或功能处接收不可信的分组流。 该特征启动了对应于不可信分组流的不可信流对象添加到流表中。 概率流管理机构确定流表中不可信流对象的数量是否低于预定的下限。 如果不可信流对象的数量超过下限，并且在将不可信流对象添加到流表中之前，概率流管理机制概率地确定是否可以将不可信流对象添加到流表中。