发明名称 System or method for selecting classifier attribute types
摘要 A system or method (collectively "selection system") is disclosed for selecting attributes (44) for a classifier (32) in a sensor system. The selection system selects attribute types (44) using statistical distributions (66) of the attribute values (46) associated with those attribute types (44). Attribute types (44) not within the selectively identified subset of attribute types (44) can be filtered out before such data is sent to a classifier (32). The system can use a test data subsystem (52) for storing and accessing actual sensor data (26). A distribution analysis subsystem (54) can perform statistical analyses on the test data to identify underlying distributions (66), and to compare individual attribute types (44) to such distributions (66). An attribute selection subsystem (56), wherein said attribute selection subsystem (56) selectively identifies a subset of attribute types (44) from said subset of attribute types (44). <IMAGE>
申请公布号 AU2004200298(A1) 申请公布日期 2004.09.16
申请号 AU20040200298 申请日期 2004.01.28
申请人 EATON CORPORATION 发明人 MICHAEL EDWARD FARMER
分类号 B60R21/16;B60R16/02;B60R16/023;B60R21/00;B60R21/01;B60R21/013;B60R21/0134;B60R21/015;G06F7/00;G06K9/00;G06K9/62;G06T7/00;G06T7/20 主分类号 B60R21/16
代理机构 代理人
主权项
地址