会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 4. 发明授权
    • Neural network element with reinforcement/attenuation learning
    • 具有加强/衰减学习的神经网络元素
    • US07664714B2
    • 2010-02-16
    • US11255895
    • 2005-10-21
    • Hiroshi TsujinoNobuaki MiyakawaGen MatsumotoRyoji Noyori, legal representative
    • Hiroshi TsujinoNobuaki MiyakawaGen Matsumoto
    • G06N3/02
    • G06N3/08G06N3/049
    • A neural network element, outputting an output signal in response to a plurality of input signals, comprises a history memory for accumulating and storing the plurality of input signals in a temporal order as history values. It also includes an output module for outputting the output signal when an internal state exceeds a predetermined threshold value, the internal state being based on a sum of the product of a plurality of input signals and corresponding coupling coefficients. The history values depend on change of the internal state. The neural network element is configured to subtract a predetermined value from the internal state immediately after the output module fires and performs learning for reinforcing or attenuating the coupling coefficient according to the history values after the output module fires.
    • 输出响应于多个输入信号的输出信号的神经网络元件包括用于以时间顺序累积并存储多个输入信号作为历史值的历史存储器。 它还包括一个输出模块,用于当内部状态超过预定阈值时输出输出信号,内部状态基于多个输入信号的乘积和相应的耦合系数之和。 历史价值取决于内部状态的变化。 神经网络元件被配置为在输出模块触发之后立即从内部状态减去预定值,并且在输出模块触发之后根据历史值执行用于加强或衰减耦合系数的学习。
    • 6. 发明授权
    • Method for assigning job in parallel processing method and parallel processing method
    • 并行处理方法和并行处理方法分配作业的方法
    • US07370328B2
    • 2008-05-06
    • US10257913
    • 2002-04-25
    • Sou YamadaShinjiro InabataNobuaki MiyakawaHajime TakashimaKunihiro KitamuraUnpei Nagashima
    • Sou YamadaShinjiro InabataNobuaki MiyakawaHajime TakashimaKunihiro KitamuraUnpei Nagashima
    • G06F9/44G06F15/16G06F15/163
    • G06F9/5066G06F9/4881
    • When parallel processing is executed by parallel computers composed of a host computer and a plurality of processors connected to the host computer through a common bus, there is provided a method of assigning jobs to respective processors with high efficiency. A job in which a ratio between a communication time and a calculation time is larger than a predetermined value or larger than a fraction of processors and a job in which a ratio between a communication time and a calculation time is smaller than a predetermined value or smaller than a fraction of processors can be alternately assigned to respective processors. Alternatively, jobs are assigned to respective processors in such a manner that a plurality of processors and a plurality of jobs are divided into a plurality of groups in a one-to-one relation, jobs in which sizes comprising communication time and calculation time and ratios between the communication times and the calculation times approximate to each other may belong to different job groups and the order in which the jobs in which the sizes comprising the communication time and the calculation time and the ratios between the communication times and the calculation times approximate to each other are assigned within respective job groups may differ from each other among a plurality of job groups.
    • 当并行处理由主计算机和通过公共总线连接到主计算机的多个处理器组成的并行计算机执行时,提供了一种以高效率将作业分配给各个处理器的方法。 其中通信时间和计算时间之间的比率大于预定值或大于处理器的一部分的工作,以及通信时间和计算时间之间的比率小于预定值或更小的作业 可以将一小部分处理器交替分配给相应的处理器。 或者,将作业分配给各个处理器,使得多个处理器和多个作业以一对一的关系被分成多个组,其中包括通信时间和计算时间和比率的尺寸的作业 在通信时间和彼此近似的计算时间之间可以属于不同的作业组,并且其中包括通信时间和计算时间的尺寸以及通信时间与计算时间之间的比率近似于 彼此分配在各个作业组中可以在多个作业组中彼此不同。
    • 8. 发明授权
    • Method and apparatus for parallel processing
    • 并行处理方法和装置
    • US06799151B1
    • 2004-09-28
    • US09544201
    • 2000-04-07
    • So YamadaShinjiro InabataNobuaki MiyakawaHajime TakashimaKunihiro KitamuraShigeru Obara
    • So YamadaShinjiro InabataNobuaki MiyakawaHajime TakashimaKunihiro KitamuraShigeru Obara
    • G06F1750
    • G06F9/5066G06F8/45G06F17/10
    • Matrix element calculation carried out efficiently without the overhead of communication between a host computer and processor elements even in parallel calculation utilizing a low-cost communication device and multiple processor elements having memories of a small capacity. In a method for calculating molecular orbitals, for example, all elements F(I, J) of a Fock matrix are calculated where an outermost loop is a loop associated with combinations (RT) of contracted shell R and contracted shell T which satisfy relationships R≦Nshell and T≦R. A second loop is a loop associated with contracted shell S, and a third loop is a loop associated with contracted shell U. Alternatively, the second loop is a loop associated with the contracted shell U, and the third loop is a loop associated with the contracted shell S. The value of S ranges from 1 to R, and the value of U ranges from 1 to R. Calculation of predetermined electron repulsion integrals and calculation of a predetermined part of Fock matrix elements utilizing the result of the calculation is carried out inside the third loop.
    • 即使在使用低成本通信装置的并行计算和具有小容量的存储器的多个处理器元件的情况下,也可以有效地进行矩阵元素计算,而无需主计算机和处理器元件之间的通信开销。 在用于计算分子轨道的方法中,例如,计算Fock矩阵的所有元素F(I,J),其中最外面的循环是与收缩壳体R和收缩壳体T的组合(RT)相关联的回路,其满足关系R <= Nshell和T <= R。 第二个循环是与收缩壳体S相关联的循环,第三个循环是与收缩的壳体U相关联的循环。或者,第二个循环是与收缩的壳体U相关联的循环,第三个循环是与 S的值范围从1到R,U的值在1到R之间。使用计算结果计算预定的电子斥力积分和计算预定部分的Fock矩阵元素 在第三个循环内。
    • 9. 发明授权
    • Coordinate difference calculating device
    • 坐标差计算装置
    • US5572447A
    • 1996-11-05
    • US160800
    • 1993-12-03
    • Shinjiro ToyodaHitoshi IkedaEiri HashimotoNobuaki Miyakawa
    • Shinjiro ToyodaHitoshi IkedaEiri HashimotoNobuaki Miyakawa
    • G06F7/544G06F17/50G01N15/02G01B21/16
    • G06F7/544G06F19/704G06F19/701
    • A device for calculating differences includes a difference circuit for generating difference signals .DELTA.x.sub.j =x.sub.j -x.sub.i, .DELTA.y.sub.j =y.sub.j -y.sub.i, and .DELTA.z.sub.j =z.sub.j -z.sub.i between coordinates of i having (x.sub.i, y.sub.i, z.sub.i) coordinate signals and coordinates of j having (x.sub.j, y.sub.j, z.sub.j) coordinate signals in an orthogonal coordinate system. The difference circuit includes an x-axis circuit, responsive to the x.sub.i and x.sub.j signals having a first circuit for receiving the x.sub.i coordinate signal and the x.sub.j coordinate signal and generating the .DELTA.x.sub.j ; a comparison circuit for comparing the x.sub.i and x.sub.j signals and determining whether the .DELTA.x.sub.j is less than a first set value -L.sub.x /2 corresponding to a length of a side of a virtual rectangular parallelepiped or greater than a second set value L.sub.x /2 corresponding to the length of the side of the virtual rectangular parallelepiped, L.sub.x being a value indicating the length of an elongated side in the x-axis direction of the virtual rectangular parallelepiped; an adder circuit for receiving the L.sub.x and .DELTA.x.sub.j and adding the L.sub.x to .DELTA.x.sub.j when .DELTA.x.sub.j is less than -L.sub.x /2; and a subtraction circuit for receiving the L.sub.x and .DELTA.x.sub.j and subtracting L.sub.x from .DELTA.x.sub.j when .DELTA.x.sub.j is greater than L.sub.x /2. The difference circuit includes y-axis and z-axis circuits similar to the x-axis circuit.
    • 用于计算差分的装置包括用于产生具有(xi,yi,zi)坐标信号和坐标的坐标之间的差分信号DELTA xj = xj-xi,DELTA yj = yj-yi和DELTA zj = zj-zi的差分电路 具有在正交坐标系中的(xj,yj,zj)坐标信号的j。 差分电路包括x轴电路,响应于具有用于接收xi坐标信号和xj坐标信号并产生DELTA xj的第一电路的xi和xj信号; 用于比较xi和xj信号并确定DELTA xj是否小于对应于虚拟长方体的一侧的长度或大于第二设定值Lx / 2的对应的第一设定值-Lx / 2的比较电路 到虚拟长方体的一侧的长度,Lx是表示虚拟长方体的x轴方向上的细长侧的长度的值; 加法器电路,用于当DELTA xj小于-Lx / 2时,接收Lx和DELTA xj并将Lx加到DELTA xj; 以及减法电路,用于当DELTA xj大于Lx / 2时,从DELTA xj接收Lx和DELTA xj并从DELTA xj中减去Lx。 差分电路包括类似于x轴电路的y轴和z轴电路。