发明名称 METHOD AND PROGRAM STRUCTURE FOR MACHINE LEARNING
摘要 A method provides program structures for constructing a program that is learned over training data. In one example, two specific program structures are provided in which the first program structure transforms each vector in an input tuple of vectors to provide an output tuple of vectors, and the second program structure operates on an input tuple of vectors to provide an output tuple of vectors by applying one or more transformations that each involves two or more vectors in the input tuple. The transformations of the first and second program structures may be linear transformations. The program may alternatively execute the first program structure and the second program structure in any suitable order a number of times, beginning with operating one of the program structures on an initial tuple of vectors. The vectors may each consist of an ordered set of real numbers.
申请公布号 US2014258206(A1) 申请公布日期 2014.09.11
申请号 US201414203306 申请日期 2014.03.10
申请人 Harik Georges 发明人 Harik Georges
分类号 G06N5/02;G06N99/00 主分类号 G06N5/02
代理机构 代理人
主权项 1. A method for constructing a program that is learned over training data, the program receiving an initial tuple of vectors in which each vector consists of an ordered set of real numbers, the method comprising: providing a first program structure, the first program structure provided for transforming each vector in an input tuple of vectors to provide an output tuple of vectors; providing a second program structure, the second program structure provided to operate on an input tuple of vectors to provide an output tuple of vectors by applying one or more transformations that each involves two or more vectors in the input tuple and by summing the results of the transformations; and alternatively executing the first program structure and the second structure a plurality of times, beginning with operating one of the program structures on the initial tuple of vectors.
地址 Palo Alto CA US