中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

179 records (took 0.048 seconds)

    SCALABLE STREAMING DECISION TREE LEARNING
    34.
    发明申请
    SCALABLE STREAMING DECISION TREE LEARNING 审中-公开
    可定量流动决策树的学习

    公开(公告)号：US20170061327A1

    公开(公告)日：2017-03-02

    申请号：US14953457

    申请日：2015-11-30

    Applicant: International Business Machines Corporation

    Inventor： Wei Shan Dong ,  Peng Gao ,  Guo Qiang Hu ,  Chang Sheng LI ,  Xu Liang Li ,  Chun Yang Ma ,  Zhi Wang ,  Xin Zhang

    IPC: G06N99/00 ,  G06N5/02

    CPC classification number: G06N5/02 ,  G06N5/025

    Abstract: In one embodiment, a computer-implemented method includes receiving training data including a plurality of records, each record having a plurality of attributes. The training data is horizontally parallelized across two or more processing elements. This horizontal parallelizing includes dividing the training data into two or more subsets of records; assigning each subset of records to a corresponding processing element of the two or more processing elements; transmitting each subset of records to its assigned processing element; and sorting, at the two or more processing elements, the two or more subsets of records to two or more candidate leaves of a decision tree. The output from horizontally parallelizing is converted into input for vertically parallelizing the training data. The training data is vertically parallelized across the two or more processing elements. The decision tree is grown based at least in part on the horizontally parallelizing, the converting, and the vertically parallelizing.

    Abstract translation: 在一个实施例中，计算机实现的方法包括接收包括多个记录的训练数据，每个记录具有多个属性。 训练数据在两个或多个处理元件之间水平并行化。 这种水平并行化包括将训练数据分成两个或更多个记录子集; 将每个记录子集分配给所述两个或多个处理元件的相应处理元件; 将每个记录子集传送到其分配的处理元件; 并且在所述两个或更多个处理元素处将所述两个或多个记录子集分类到决策树的两个或更多个候选叶。 水平并行化的输出被转换为用于垂直并行化训练数据的输入。 训练数据跨越两个或多个处理元件垂直并行化。 决策树至少部分地基于水平并行化，转换和垂直并行化生长。

    SCALABLE STREAMING DECISION TREE LEARNING
    35.
    发明申请
    SCALABLE STREAMING DECISION TREE LEARNING 审中-公开

    公开(公告)号：US20170061318A1

    公开(公告)日：2017-03-02

    申请号：US14833397

    申请日：2015-08-24

    Applicant: International Business Machines Corporation

    Inventor： Wei Shan Dong ,  Peng Gao ,  Guo Qiang Hu ,  Cheng Sheng LI ,  Xu Liang Li ,  Chun Yang Ma ,  Zhi Wang ,  Xin Zhang

    IPC: G06N99/00

    Abstract: In one embodiment, a computer-implemented method includes receiving training data including a plurality of records, each record having a plurality of attributes. The training data is horizontally parallelized across two or more processing elements. This horizontal parallelizing includes dividing the training data into two or more subsets of records; assigning each subset of records to a corresponding processing element of the two or more processing elements; transmitting each subset of records to its assigned processing element; and sorting, at the two or more processing elements, the two or more subsets of records to two or more candidate leaves of a decision tree. The output from horizontally parallelizing is converted into input for vertically parallelizing the training data. The training data is vertically parallelized across the two or more processing elements. The decision tree is grown based at least in part on the horizontally parallelizing, the converting, and the vertically parallelizing.

    Vehicle domain multi-level parallel buffering and context-based streaming data pre-processing system
    36.
    发明授权
    Vehicle domain multi-level parallel buffering and context-based streaming data pre-processing system 有权
    车载多级并行缓冲和基于上下文的流数据预处理系统

    公开(公告)号：US09537914B1

    公开(公告)日：2017-01-03

    申请号：US14955218

    申请日：2015-12-01

    Applicant: International Business Machines Corporation

    Inventor： Wei Shan Dong ,  Ning Duan ,  Peng Gao ,  Chang Sheng Li ,  Chun Yang Ma ,  Xin Zhang ,  Jun Zhu

    IPC: G06F15/16 ,  H04L29/06 ,  H04L12/861

    CPC classification number: H04L65/4069 ,  G06F9/4881 ,  G08G1/0112 ,  H04L67/12

    Abstract: A vehicle domain multi-level parallel buffering and context-based streaming data pre-processing system includes a first data processing level and a second data processing level. The first data processing level includes a first-level buffer configured to buffer data provided from a plurality of raw data streams output from a plurality of vehicles. The second data processing level includes an electronic task-queue-dictionary (TQD) module and a plurality of second-level data processing buffers. The TQD module is configured to create a plurality of tasks in response to receiving a serial data stream output from the first-level buffer. The TQD module is further configured to assign each task to a corresponding second-level buffer, and separate the serial data stream into individual data values that are delivered to a specific second-level buffer based on the task so as to generate a multi-level parallel context-based buffering operation.

    Abstract translation: 车载域多级并行缓冲和基于上下文的流数据预处理系统包括第一数据处理级和第二数据处理级。 第一数据处理级包括被配置为缓冲从多个车辆输出的多个原始数据流提供的数据的第一级缓冲器。 第二数据处理级包括电子任务队列字典（TQD）模块和多个第二级数据处理缓冲器。 TQD模块被配置为响应于接收从第一级缓冲器输出的串行数据流而创建多个任务。 TQD模块还被配置为将每个任务分配给相应的二级缓冲器，并且将串行数据流分离成基于该任务传送到特定二级缓冲器的单独数据值，以便生成多级 并行上下文缓冲操作。

    Method and apparatus of road location inference for moving object
    37.
    发明授权
    Method and apparatus of road location inference for moving object 有权
    运动物体道路位置推理方法与装置

    公开(公告)号：US09494694B1

    公开(公告)日：2016-11-15

    申请号：US14963409

    申请日：2015-12-09

    Applicant: International Business Machines Corporation

    Inventor： Wei Shan Dong ,  Ning Duan ,  Peng Gao ,  Chun Yang Ma ,  Zhi Hu Wang ,  Xin Zhang

    IPC: G01S19/42 ,  G01S19/43 ,  G01S19/52

    CPC classification number: G01S19/50 ,  G01C21/30 ,  G01S19/39

    Abstract: A method and apparatus of location sequence inferences for moving objects traveling along a path. The method and apparatus primarily concerns determining the location of a moving vehicle on a roadway in a roadway network. The inputs to the system include: raw GPS tracking sequence with timestamp, trajectory of the moving object inferred by map matching, accurate speed sequence from a reliable device, e.g. OBD (On-Board Diagnostics is an automotive term referring to a vehicle's self-diagnostic and reporting capability), historical map matching results and historical locations sequence inference results. The output of the system is a sequence of more accurate location (on road segments) sequences than raw GPS locations and map matching results.

    Abstract translation: 一种用于沿着路径行进的移动物体的位置序列推断的方法和装置。 该方法和设备主要涉及确定移动车辆在道路网中的道路上的位置。 系统的输入包括：具有时间戳的原始GPS跟踪序列，通过地图匹配推断的移动物体的轨迹，来自可靠设备的准确速度序列，例如。 OBD（车载诊断是汽车术语，指车辆的自我诊断和报告功能），历史地图匹配结果和历史位置序列推断结果。 系统的输出是比原始GPS位置更精确的位置（在路段上）序列的序列，并且映射匹配结果。

    VERIFYING A ROAD NETWORK OF A MAP
    38.
    发明申请
    VERIFYING A ROAD NETWORK OF A MAP 有权

    公开(公告)号：US20160298973A1

    公开(公告)日：2016-10-13

    申请号：US15185085

    申请日：2016-06-17

    Applicant: International Business Machines Corporation

    Inventor： Ning Duan ,  Peng Gao ,  Baohua Wang ,  Xin Zhang

    IPC: G01C21/32 ,  G01C21/34

    CPC classification number: G01C21/32 ,  G01C21/20 ,  G01C21/26 ,  G01C21/30 ,  G01C21/34 ,  G01C21/3667

    Abstract: The present disclosure relates to a computer implemented method, a computer program product, and a computer system for verifying the road network of map. An embodiment of the present invention provides a method for verifying a road network of a map. The method comprises: determining, based on locations of a set of points in the map, a trajectory distance of a trajectory between a start point and an end point in the set of points on the map; determining a route distance between the start point and the end point in the road network of the map; and verifying the road network by comparing the trajectory distance and the route distance.

    GENERATING A QUERY INDEX AND QUERYING ON THE BASIS OF THE QUERY INDEX
    39.
    发明申请
    GENERATING A QUERY INDEX AND QUERYING ON THE BASIS OF THE QUERY INDEX 有权
    根据查询索引生成查询索引和查询

    公开(公告)号：US20160033298A1

    公开(公告)日：2016-02-04

    申请号：US14807923

    申请日：2015-07-24

    Applicant: International Business Machines Corporation

    Inventor： Ning Duan ,  Peng Gao ,  Baohua Wang ,  Feng Juan Wang ,  Xin Zhang

    IPC: G01C21/36 ,  G08G1/01 ,  G01C21/34

    CPC classification number: G01C21/3691 ,  G01C21/3415 ,  G08G1/0112 ,  G08G1/0125 ,  G08G1/0133

    Abstract: A road segment set influenced by an event on the basis of a connected network is identified. The connected network is built on the basis of reachability of a road segment in the road network. Furthermore, each road segment in the road segment set is reachable via the road network from the location of the event, or, in the alternative, the location of the event is reachable via the road network from each road segment in the road segment set. A query index is generated on the basis of the event and the road segment set.

    Abstract translation: 识别由基于连接网络的事件影响的路段集合。 连接网络建立在道路网路段可达性的基础上。 此外，路段集合中的每个路段可以通过道路网从事件的位置到达，或者替代地，事件的位置可以经由道路网从道路段集合中的每个路段到达。 基于事件和路段集合生成查询索引。

    IDENTIFYING SUBSURFACE MATERIAL LAYER
    40.
    发明申请
    IDENTIFYING SUBSURFACE MATERIAL LAYER 审中-公开
    识别表面材料层

    公开(公告)号：US20150120195A1

    公开(公告)日：2015-04-30

    申请号：US14527272

    申请日：2014-10-29

    Applicant: International Business Machines Corporation

    Inventor： Wei S. Dong ,  Wen Q. Huang ,  Chun Y. Ma ,  Chunhua Tian ,  Yu Wang ,  Junchi Yan ,  Chao Zhang ,  Xin Zhang

    IPC: G01V99/00 ,  G06N99/00

    CPC classification number: G01V99/00 ,  G01V2210/62 ,  G01V2210/66

    Abstract: In an approach for identifying subsurface material layers, a computer processor: acquires a well log of a location to be explored, the log comprising data corresponding to multiple geophysical parameters, the data of each parameter comprising measurement values of the parameter at different depths underground; matches a reference data of each parameter corresponding to each of multiple layer transition types with the data of that parameter in the well log at depths underground, wherein each layer transition type indicates an upper material layer and a lower material layer, and the reference data is used to represent a variation trend of the parameter in a transitional zone conforming with the layer transition type; and according to the matching result, determines a layer transition type at the location to be explored and a depth of an interface between the upper material layer and the lower indicated by the layer transition type.

    Abstract translation: 在一种用于识别地下物质层的方法中，计算机处理器：获取要探索的位置的测井记录，所述对数包括对应于多个地球物理参数的数据，每个参数的数据包括在地下不同深度的参数的测量值; 将与多层转移类型中的每一个对应的每个参数的参考数据与地下深井中的该对数据中的该参数的数据进行匹配，其中每个层转换类型表示上层材料层和下层材料层，参考数据为 用于表示符合层过渡类型的过渡区域中参数的变化趋势; 并且根据匹配结果确定要探索的位置处的层转变类型以及由层过渡类型指示的上层材料层和下层之间的界面的深度。

Patent Agency Ranking