利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

6 条记录 (耗时 0.017 秒)

    Method for determination of weights, suitable for elimination, of a neural network using a computer
    1.
    发明授权
    Method for determination of weights, suitable for elimination, of a neural network using a computer 有权
    使用计算机确定适合于消除神经网络的权重的方法

    公开(公告)号：US06311172B1

    公开(公告)日：2001-10-30

    申请号：US09155197

    申请日：1998-09-23

    申请人： Volker Tresp ,  Hans-Georg Zimmermann ,  Ralph Neuneier

    发明人： Volker Tresp ,  Hans-Georg Zimmermann ,  Ralph Neuneier

    IPC分类号： G06F1518

    CPC分类号： G06N3/08

    摘要： The training phase of a neural network NN is stopped before an error function, which is to be minimized in the training phase, reaches a minimum (301). A first variable (EG) is defined using, for example, the optimal brain damage method or the optimal brain surgeon method, on the assumption that the error function is at the minimum. Furthermore, a second variable (ZG) is determined which provides an indication of the manner in which the value of the error function varies when a weight (wi) is removed from the neural network (NN). The first variable (EG) and the second variable (ZG) are used to classify the weight (wi) as being suitable or unsuitable for removal from the neural network (NN).

    摘要翻译： 在训练阶段要最小化的误差函数达到最小（301）之前，停止神经网络NN的训练阶段。 假设误差函数为最小，使用例如最佳脑损伤方法或最佳脑外科医生方法来定义第一变量（EG）。 此外，确定第二变量（ZG），当从神经网络（NN）移除权重（wi）时，提供误差函数的值变化的方式的指示。 第一变量（EG）和第二变量（ZG）用于将权重（wi）分类为适合于或不适合于从神经网络（NN）中移除。

    Method of training a neural network
    2.
    发明授权
    Method of training a neural network 有权
    训练神经网络的方法

    公开(公告)号：US06247001B1

    公开(公告)日：2001-06-12

    申请号：US09142275

    申请日：1998-09-03

    申请人： Volker Tresp ,  Ralph Neuneier

    发明人： Volker Tresp ,  Ralph Neuneier

    IPC分类号： G06F9445

    CPC分类号： G06N3/08

    摘要： A state vector (SVt) is determined with elements that characterize a financial market (101). Taking into account predetermined evaluation variables, an evaluation (Vt) is determined (102) for the state vector (SVt). In addition, a chronologically following state vector (SVt+1) is determined (103) and evaluated (Vt+1). On the basis of the two evaluations (Vt, Vt+1), weights (wi) of the neural network (NN) are adapted (104) using a reinforcement learning method (&Dgr;wi).

    摘要翻译： 状态向量（SVt）由元素表征金融市场（101）确定。 考虑到预定的评估变量，对于状态向量（SVt）确定评估（Vt）（102）。 另外，确定（103）并计算（Vt + 1）的时间顺序的跟随状态向量（SVt + 1）。 在两个评估（Vt，Vt + 1）的基础上，使用强化学习方法（DELTAwi）来适应神经网络（NN）的权重（wi）（104）。

    Method for detecting discrepancies between a user's perception of web sites and an author's intention of these web sites
    3.
    发明申请
    Method for detecting discrepancies between a user's perception of web sites and an author's intention of these web sites 审中-公开
    用于检测用户对网站的感知与作者对这些网站的意图之间的差异的方法

    公开(公告)号：US20070088720A1

    公开(公告)日：2007-04-19

    申请号：US11250573

    申请日：2005-10-17

    申请人： Ralph Neuneier ,  Michal Skubacz ,  Carsten Stolz ,  Maximilian Vermetz

    发明人： Ralph Neuneier ,  Michal Skubacz ,  Carsten Stolz ,  Maximilian Vermetz

    IPC分类号： G06F7/00

    CPC分类号： G06F16/957

    摘要： Method of computer-based detection of discrepancy between a user's perception of web sites and an author's intention of these web sites, wherein user interactions are gathered and combined with the content of individual web pages, the combination thereof is clustered topically, and a respective topical distance of the web pages is compared to a structural distance of the web pages, which results from the author's elected arrangement of the web pages to each other, whereby the difference in both distances gives the discrepancy in the user's perception and the author's intention of the web pages, characterized in that at least parts of the text are extracted from the web pages for building keywords, which represent the contents of such web pages.

    摘要翻译： 基于计算机的检测用户对网站的感知与作者对这些网站的意图之间的差异的方法，其中用户交互被收集并与各个网页的内容相结合，其组合是局部聚类的，并且各自的主题 将网页的距离与网页的结构距离进行比较，网页的结构距离是由作者选择的网页排列而导致的，因此两个距离的差异给出了用户感知的差异，以及作者的意图 网页，其特征在于，从网页中提取文本的至少部分，以构建表示这些网页的内容的关键字。

    Communication system for processing data
    6.
    发明授权
    Communication system for processing data 失效
    通讯系统处理数据

    公开(公告)号：US08700647B2

    公开(公告)日：2014-04-15

    申请号：US11763008

    申请日：2007-06-14

    申请人： Ralph Neuneier ,  Michal Skubacz

    发明人： Ralph Neuneier ,  Michal Skubacz

    IPC分类号： G06F17/30

    CPC分类号： H04L67/02

    摘要： In a plurality of sub-systems, a respective sub-system processes data according to a logic system specific to the respective sub-system. At least some of the sub-systems respectively include at least one acquisition unit which is configured in such a way that it acquires information relating to the processed data of the respective sub-system and converts it into a predefined data structure which is the same for at least some of the sub-systems. the acquisition units are connected to a common analysis unit which during operation of the communication system receives information from the acquisition units relating to the processed data in the predefined data structure and analyzes it in relation to predefined criteria.

    摘要翻译： 在多个子系统中，相应的子系统根据对各个子系统特定的逻辑系统来处理数据。 至少一些子系统分别包括至少一个采集单元，其被配置为使得其获取与相应子系统的处理数据相关的信息，并将其转换为对于 至少一些子系统。 采集单元连接到公共分析单元，其在通信系统的操作期间从与预定数据结构中的处理数据相关的获取单元接收信息并且相对于预定标准进行分析。