公开(公告)号：US06823324B2

公开(公告)日：2004-11-23

申请号：US09838765

申请日：2001-04-19

申请人： Hideki Kagemoto ,  Takashi Kakiuchi ,  Toshihiro Hishida ,  Kazuo Okamura

发明人： Hideki Kagemoto ,  Takashi Kakiuchi ,  Toshihiro Hishida ,  Kazuo Okamura

IPC分类号： G06F1518

CPC分类号： H04N21/26291 ,  H04N7/165 ,  H04N21/8153 ,  H04N21/854

摘要： A template storage unit stores a plurality of templates. A material data receiving unit receives material data. A combining rule storage unit stores a plurality of combining rules that each indicate a combining condition as a combination of attribute information, such as a type of material data and a version number, and a name of a template to be combined with when the combining condition is met. A combining instruction unit judges whether a combining condition of any of the combining rules is met or not and instructs a combining unit to combine the material data with the template when the combining condition is met. The combining unit combines the material data or a combination of the material data with the specified template according to the combining rule.

摘要翻译： 模板存储单元存储多个模板。 材料数据接收单元接收材料数据。 组合规则存储单元存储多个组合规则，每个组合规则指示组合条件作为诸如素材数据类型和版本号的属性信息的组合，以及当组合条件 遇到了 组合指令单元判断是否满足任何组合规则的组合条件，并指示组合单元在满足组合条件时将材料数据与模板组合。 组合单元根据组合规则将材料数据或材料数据的组合与指定的模板相结合。

公开(公告)号：US06810391B1

公开(公告)日：2004-10-26

申请号：US09388627

申请日：1999-09-02

申请人： Thomas Birkhoelzer ,  Joachim Horn ,  Marco Pellegrino

发明人： Thomas Birkhoelzer ,  Joachim Horn ,  Marco Pellegrino

IPC分类号： G06F1518

CPC分类号： A61B5/0002 ,  A61B5/7264 ,  G06F19/00 ,  G16H10/20 ,  G16H15/00 ,  G16H40/40 ,  G16H40/63 ,  Y10S128/904 ,  Y10S128/924 ,  Y10S706/924

摘要： a System for calculating and emitting analytical data, particularly medically relevant analytical data, has a computer with an evaluation unit for analyzing information entered by the system user and for generating the analytical data, and an output medium provided at the user side. The evaluation unit is configured for interrogating and accepting information that can be entered at the user side such that the information can be entered autonomously by the user based on a number of inquiries which are predefined at the system side, and which can be selected with reference to each response, and which can be made available as an output to user via the output medium. The evaluation unit is configured for controlling an interactive information capture such that, depending on the running information compilation, specific inquiries can be formulated and shown to the user. The evaluation unit is configured for processing information that is random in content and for evaluating the analytical data with a value measure lying between two extreme values. The analytical data that is generated on the basis of the running information compilation can be interrogated by the user at any time and supplied to the user as an output upon such interrogation.

摘要翻译： 用于计算和发布分析数据的系统，特别是医学上相关的分析数据，具有计算机，其具有用于分析由系统用户输入的信息和生成分析数据的评估单元，以及在用户侧提供的输出介质。 评估单元被配置为用于询问和接受可以在用户侧输入的信息，使得用户可以基于在系统侧预定义的多个查询来自主地输入信息，并且可以通过参考来选择该信息 到每个响应，并且哪些可以通过输出介质作为输出提供给用户。 评估单元被配置为控制交互式信息捕获，使得根据运行信息编译，可以为用户制定并显示特定的查询。 评估单元被配置用于处理在内容中是随机的信息，并用用于在两个极值之间的值测量来评估分析数据。 基于运行信息编译生成的分析数据可以由用户在任何时间询问，并且在这种询问时作为输出提供给用户。

公开(公告)号：US06763338B2

公开(公告)日：2004-07-13

申请号：US10116835

申请日：2002-04-05

申请人： Evan R. Kirshenbaum

发明人： Evan R. Kirshenbaum

IPC分类号： G06F1518

CPC分类号： G06K9/6292 ,  G06N3/086

摘要： A method and apparatus for computing an overall or aggregate decision based on intermediate decisions as to which of a set of alternatives best characterize an object. The alternatives are partitioned into at least two series of preferences corresponding to at least two intermediate rankings. Various embodiments may base the intermediate rankings on: a machine learning technique; a decision tree; a belief network; a neural network; a static model; a program; or an evolutionary training method. Based on the preferences, a weak alternative is selected and removed from the series. The selection of the weak alternative may include identifying which alternatives lose pairwise to the other alternatives, are excluded from the first preferences, are included in the last preferences, or have a lowest average preference ranking. The selecting and removing continue until the remaining alternatives are the aggregate decision. Various embodiments may be applied to classification problems, prediction problems or selection problems.

摘要翻译： 一种用于基于关于一组替代物中的哪一种最佳地表征对象的中间决策来计算总体或聚合决策的方法和装置。 备选方案被划分为对应于至少两个中间排名的至少两个系列的偏好。 各种实施例可以基于以下机器学习技术的中间排名： 决策树 信仰网络; 神经网络 静态模型; 一个程序 或进化训练方法。 根据偏好，从系列中选出并删除了弱选项。 弱选择的选择可以包括确定哪些替代物与其他替代方案丢失成对，从第一偏好中排除，包括在最后的偏好中，或者具有最低的平均偏好排名。 选择和删除继续，直到剩余的替代方案是总决定。 各种实施例可以应用于分类问题，预测问题或选择问题。

公开(公告)号：US06760714B1

公开(公告)日：2004-07-06

申请号：US09675867

申请日：2000-09-29

申请人： William R. Caid ,  Robert Hecht-Neilsen

发明人： William R. Caid ,  Robert Hecht-Neilsen

IPC分类号： G06F1518

CPC分类号： G06F17/30265 ,  G06F17/30256 ,  G06F17/3061 ,  G06K9/4623

摘要： Image features are generated by performing wavelet transformations at sample points on images stored in electronic form. Multiple wavelet transformations at a point are combined to form an image feature vector. A prototypical set of feature vectors, or atoms, is derived from the set of feature vectors to form an “atomic vocabulary.” The prototypical feature vectors are derived using a vector quantization method (e.g., using neural network self-organization techniques) in which a vector quantization network is also generated. The atomic vocabulary is used to define new images. Meaning is established between atoms in the atomic vocabulary. High-dimensional context vectors are assigned to each atom. The context vectors are then trained as a function of the proximity and co-occurrence of each atom to other atoms in the image. After training, the context vectors associated with the atoms that comprise an image are combined to form a summary vector for the image. Images are retrieved using a number of query methods (e.g., images, image portions, vocabulary atoms, index terms). The user's query is converted into a query context vector. A dot product is calculated between the query vector and the summary vectors to locate images having the closest meaning. The invention is also applicable to video or temporally related images, and can also be used in conjunction with other context vector data domains such as text or audio, thereby linking images to such data domains.

摘要翻译： 通过在以电子形式存储的图像上的采样点执行小波变换来生成图像特征。 将点处的多个小波变换组合以形成图像特征向量。 特征向量或原子的原型集是从特征向量集合中导出的，以形成“原子词汇”。 使用其中也生成矢量量化网络的矢量量化方法（例如，使用神经网络自组织技术）导出原型特征向量。 原子词汇用于定义新图像。 在原子词汇中的原子之间建立意义。 高维上下文向量分配给每个原子。 然后将上下文矢量作为每个原子与图像中其他原子的邻近和共现的函数进行训练。 在训练之后，与构成图像的原子相关联的上下文向量被组合以形成图像的汇总向量。 使用许多查询方法（例如，图像，图像部分，词汇原子，索引项）来检索图像。 用户的查询被转换为查询上下文向量。 在查询向量和汇总向量之间计算点积，以定位具有最接近意义的图像。 本发明也适用于视频或时间相关的图像，并且还可以与诸如文本或音频的其他上下文矢量数据域一起使用，从而将图像链接到这样的数据域。

公开(公告)号：US06748369B2

公开(公告)日：2004-06-08

申请号：US09337285

申请日：1999-06-21

申请人： Pratap Shankar Khedkar ,  Piero Patrone Bonissone

发明人： Pratap Shankar Khedkar ,  Piero Patrone Bonissone

IPC分类号： G06F1518

CPC分类号： G06N5/048 ,  G06N3/0436

摘要： A method and system for automating a process for valuing a property that produces an estimated value of a subject property, and a reliability assessment of the estimated value. The process is a generative artificial intelligence method that trains a fuzzy-neural network using a subset of cases from a case-base, and produces a run-time system to provide an estimate of the subject property's value. A network-based implementation of fuzzy inference is based on a system that implements a fuzzy system as a five-layer neural network so that the structure of the network can be interpreted in terms of high-level rules. The neural network is trained automatically from data. IF/THEN rules are used to map inputs to outputs by a fuzzy logic inference system. Different models for the same problem can be obtained by changing the inputs to the neuro-fuzzy network, or by varying its architecture.

摘要翻译： 用于自动化产生对象属性的估计值的属性的评估过程的方法和系统以及估计值的可靠性评估。 该过程是一种生成人工智能方法，它使用案例基础案例的子集训练模糊神经网络，并产生一个运行时系统来提供主体属性值的估计。 模糊推理的基于网络的实现是基于将模糊系统实现为五层神经网络的系统，以便可以根据高级规则来解释网络的结构。 从数据自动训练神经网络。 IF / THEN规则用于通过模糊逻辑推理系统将输入映射到输出。 通过改变对神经 - 模糊网络的输入，或通过改变其架构，可以获得同样问题的不同模型。

公开(公告)号：US06741973B1

公开(公告)日：2004-05-25

申请号：US08950230

申请日：1997-10-14

申请人： Lee G. Dove ,  Graham I. Johnson

发明人： Lee G. Dove ,  Graham I. Johnson

IPC分类号： G06F1518

CPC分类号： G06N3/126

摘要： A model of consumer behavior in a transaction environment such as customers moving around a bank branch, is generated from an artificial life algorithm to create a number of agents. In each agent, a genetically encoded drive, equivalent for example to hunger, is defined so as to correspond to a transaction need such as the need for cash. Interaction rules, such as navigation rules, are set for interaction between the agents and a first representation of an environment, and the program is run and the agents observed, then compared with real human behavior. The best matched agents are selected and the program run again, the steps being repeated until a required level of comparison with real behavior is reached. The model can then be used with different transaction environments to study customer behavior and to select the best branch layout or the like.

摘要翻译： 在诸如客户围绕银行分行移动的交易环境中的消费者行为的模型是从人为生命算法生成的，以创建多个代理。 在每个代理中，基因编码的驱动器（例如与饥饿等价）被定义为对应于诸如需要现金的交易需求。 交互规则（如导航规则）被设置为代理之间的交互和环境的第一表示，并且运行程序并观察代理，然后与真实的人类行为进行比较。 选择最匹配的代理程序，程序再次运行，重复这些步骤，直到达到与实际行为的比较所需的级别。 然后，该模型可以与不同的交易环境一起使用，以研究客户行为并选择最佳分支布局等。

公开(公告)号：US06738753B1

公开(公告)日：2004-05-18

申请号：US09642356

申请日：2000-08-21

申请人： Michael Andrew Hogan

发明人： Michael Andrew Hogan

IPC分类号： G06F1518

CPC分类号： G06N5/043

摘要： A modular artificial intelligence learning entity (a “golem”) which is replicated many times to form a super-entity that shows intelligent behavior transcending that of its individual constituents. Within the group of golems, individual golems may occupy roles, and are role differentiated, in that structurally identical entities perform different functions and exhibit different behavior depending on their personas and the learning they have completed as driven by other entities. The group of golems is hierarchically organized, in the sense that ‘superior’ entities issue policies to ‘subordinate’ entities. In addition to responding to ‘sense’ input from its environment, the golem responds to policy requirements set by other entities, including its superiors, and in turn sets policy requirements for its subordinates. Actions of the golem are measured for successful compliance with that golem's policies by its superior, who directs the golem's learning process. The super-entity thus gains intelligence through the policy reinforcement occurring in each superior-subordinate relationship. This scheme is well adapted to working over a network with logically separated but communicating golems. Its flexibility allows its application both to single complex problems and to repetitively occurring simple problems. Opportunities for its use arise in operating environments, in simulation and gaming, and in research.

摘要翻译： 模仿人工智能学习实体（“golem”）被复制了许多次，形成了一个超实体，超越其个人成分的智慧行为。 在群体中，个别的角色可能会扮演角色，并且在角色上有区别，在结构上相同的实体中，根据他们的角色和其他实体驱动的学习，执行不同的功能并表现出不同的行为。 在这个意义上，“上级”实体向“下属”实体发出政策，这种群体的组织是分级组织的。 除了响应其环境的“意义”输入外，灵魂还响应了其他实体（包括其上级）设定的政策要求，从而为其下属制定政策要求。 衡量灵魂的行动是衡量其上级成功遵守该公司政策的指导，他们指导了公司的学习过程。 因此，超级实体通过每个上下级关系中发生的政策强化获得智力。 该方案适用于通过逻辑分离但通信的网络进行网络工作。 其灵活性允许其应用于单个复杂问题和重复出现的简单问题。 在运营环境，模拟和游戏以及研究中都会出现机会。

公开(公告)号：US06735577B2

公开(公告)日：2004-05-11

申请号：US09898755

申请日：2001-07-02

申请人： Horst Zuse

发明人： Horst Zuse

IPC分类号： G06F1518

CPC分类号： G06F17/30277

摘要： An apparatus is configured to automatically search for relevant picture data sets from a quantity of n (n≧2) picture data sets electronically stored in a memory device. Picture attributes for each of the n picture data sets is stored electronically in the memory device, and the n picture data sets as well as the stored picture attributes are adapted to be processed electronically by processor device. A training quantity including the electronically stored picture attributes of picture data sets which were selected by a user is utilized in the context of a machine learning process carried out automatically with the aid of the processor device to determine a decision function f. Subsequently, system relevancies of the picture data sets are calculated, based on the decision function f. Selected picture data sets then are provided in the context of the search result to be output in response to the respective system relevance.

摘要翻译： 一种装置被配置为从电子存储在存储装置中的n（n> = 2）个图像数据集的数量自动搜索相关图像数据集。 n个图像数据组中的每一个的图像属性以电子方式存储在存储器件中，并且n个图像数据集以及所存储的图像属性适于由处理器装置电子处理。 包括由用户选择的图像数据集的电子存储图像属性的训练量在利用处理器装置自动执行的机器学习过程的上下文中被利用，以确定判定函数f。 随后，基于决策函数f计算图像数据集的系统相关性。 所选择的图像数据集然后在搜索结果的上下文中被提供以响应于相应的系统相关性被输出。

公开(公告)号：US06721718B2

公开(公告)日：2004-04-13

申请号：US10190322

申请日：2002-07-02

申请人： Sergei V. Ulyanov

发明人： Sergei V. Ulyanov

IPC分类号： G06F1518

CPC分类号： G05B13/0265 ,  G05B13/0285 ,  Y10S706/903

摘要： A reduced control system suitable for control of a nonlinear or unstable plant is described. The reduced control system is configured to use a reduced sensor set for controlling the plant without significant loss of control quality (accuracy) as compared to an optimal control system with an optimum sensor set. The control system calculates the information content provided by the reduced sensor set as compared to the information content provided by the optimum set. The control system also calculates the difference between the entropy production rate of the plant and the entropy production rate of the controller. A genetic optimizer is used to tune a fuzzy neural network in the reduced controller. A fitness function for the genetic optimizer provides optimum control accuracy in the reduced control system by minimizing the difference in entropy production while maximizing the sensor information content.

摘要翻译： 描述了一种适用于控制非线性或不稳定植物的减少控制系统。 与具有最佳传感器组的最佳控制系统相比，减小的控制系统被配置为使用减小的传感器组来控制设备而不显着损失控制质量（精度）。 与由最佳集合提供的信息内容相比，控制系统计算由缩小传感器组提供的信息内容。 控制系统还计算了工厂熵产生率与控制器熵产生率之间的差异。 遗传优化器用于调节减少控制器中的模糊神经网络。 遗传优化器的适应度函数通过最小化熵产生的差异，同时最大化传感器信息内容，在减少的控制系统中提供最佳的控制精度。

公开(公告)号：US06718315B1

公开(公告)日：2004-04-06

申请号：US09740067

申请日：2000-12-18

申请人： Christopher A. Meek ,  David M. Chickering ,  Jeffrey R. Bernhardt ,  Robert L. Rounthwaite

发明人： Christopher A. Meek ,  David M. Chickering ,  Jeffrey R. Bernhardt ,  Robert L. Rounthwaite

IPC分类号： G06F1518

CPC分类号： G06N99/005

摘要： Disclosed is a system for approximating conditional probabilities using an annotated decision tree where predictor values that did not exist in training data for the system are tracked, stored, and referenced to determine if statistical aggregation should be invoked. Further disclosed is a system for storing statistics for deriving a non-leaf probability corresponding to predictor values, and a system for aggregating such statistics to approximate conditional probabilities.

摘要翻译： 公开了一种使用注释决策树近似条件概率的系统，其中跟踪，存储和引用系统的训练数据中不存在的预测值，以确定是否应调用统计聚合。 进一步披露的是用于存储用于导出与预测值相对应的非叶概率的统计的系统，以及用于将这种统计量聚合以近似条件概率的系统。