公开(公告)号：US20050251496A1

公开(公告)日：2005-11-10

申请号：US11061974

申请日：2005-02-17

Applicant: Dennis DeCoste ,  Gary Flake ,  Peter Savich

Inventor： Dennis DeCoste ,  Gary Flake ,  Peter Savich

IPC: G06F17/30 ,  G06F7/00

CPC classification number: G06F17/30707 ,  G06F17/30864 ,  G06F17/30882 ,  G06F17/30991 ,  G06Q30/02

Abstract: A method for providing search results to a user is disclosed. The method includes receiving a first set of information associated with a plurality of web pages. A second set of information associated with a user preference, determining a commercial score for each web page is also received. A subset of the first set of information is determined based on the second set of information. A visual indicator for the subset of the first set of information is generated in accordance with a commercial score, and the subset and the visual indicator are displayed on a display.

Abstract translation: 公开了一种向用户提供搜索结果的方法。 该方法包括接收与多个网页相关联的第一组信息。 还接收与用户偏好相关联的第二组信息，确定每个网页的商业得分。 基于第二组信息来确定第一组信息的子集。 根据商业评分生成第一组信息的子集的视觉指示符，并且将子集和视觉指示符显示在显示器上。

公开(公告)号：US20120323682A1

公开(公告)日：2012-12-20

申请号：US13524187

申请日：2012-06-15

Applicant: Amit Umesh Shanbhag ,  Nausher Ahmed Cholavaram ,  Senthil Kumar Pandurangan ,  Zeqian Shen ,  Dennis Decoste

Inventor： Amit Umesh Shanbhag ,  Nausher Ahmed Cholavaram ,  Senthil Kumar Pandurangan ,  Zeqian Shen ,  Dennis Decoste

IPC: G06Q30/02

CPC classification number: G06Q30/0641 ,  G06Q30/0601

Abstract: Systems and methods for behavioral modeling to optimize shopping cart conversion are discussed. For example, a method can include identifying a user interacting with a networked system, accessing user profile data associated with the user, tracking user activity associated with the user, accessing a behavioral model, applying the behavioral model, and determining a shopping cart optimization. The behavioral model can be generated from historical data detailing interactions with the networked system. The behavioral model can be applied to the user profiled data and the user activity data to assist in selection of a shopping cart optimization.

Abstract translation: 讨论了用于优化购物车转换的行为建模的系统和方法。 例如，方法可以包括识别与网络系统交互的用户，访问与用户相关联的用户简档数据，跟踪与用户相关联的用户活动，访问行为模型，应用行为模型以及确定购物车优化。 行为模型可以从详细描述与网络系统的交互的历史数据生成。 行为模型可以应用于用户分析数据和用户活动数据，以帮助选择购物车优化。

公开(公告)号：US20160117587A1

公开(公告)日：2016-04-28

申请号：US14582059

申请日：2014-12-23

Applicant: Zhicheng Yan ,  Robinson Piramuthu ,  Vignesh Jagadeesh ,  Wei Di ,  Dennis Decoste

Inventor： Zhicheng Yan ,  Robinson Piramuthu ,  Vignesh Jagadeesh ,  Wei Di ,  Dennis Decoste

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06N3/04 ,  G06N3/0454 ,  G06N3/084

Abstract: Hierarchical branching deep convolutional neural networks (HD-CNNs) improve existing convolutional neural network (CNN) technology. In a HD-CNN, classes that can be easily distinguished are classified in a higher layer coarse category CNN, while the most difficult classifications are done on lower layer fine category CNNs. Multinomial logistic loss and a novel temporal sparsity penalty may be used in HD-CNN training. The use of multinomial logistic loss and a temporal sparsity penalty causes each branching component to deal with distinct subsets of categories.

Abstract translation: 分层分支深卷积神经网络（HD-CNN）改进了现有的卷积神经网络（CNN）技术。 在HD-CNN中，可以容易地区分的类分类为较高层粗类CNN，而最难分类则在下层精细类别CNN上完成。 多项式物流损失和新颖的时间稀疏性惩罚可用于HD-CNN训练。 使用多项物流损失和时间稀疏惩罚导致每个分支组件处理不同的类别子集。

公开(公告)号：US20070011110A1

公开(公告)日：2007-01-11

申请号：US11432764

申请日：2006-05-10

Applicant: Sathiya Selvaraj ,  Dennis DeCoste

Inventor： Sathiya Selvaraj ,  Dennis DeCoste

IPC: G06F15/18

CPC classification number: G06K9/6269

Abstract: Support vector machines (SVMs), though accurate, are not preferred in applications requiring great classification speed, due to the number of support vectors being large. To overcome this problem a primal system and method with the following properties has been devised: (1) it decouples the idea of basis functions from the concept of support vectors; (2) it greedily finds a set of kernel basis functions of a specified maximum size (dmax) to approximate the SVM primal cost function well; (3) it is efficient and roughly scales as O(ndmax2) where n is the number of training examples; and, (4) the number of basis functions it requires to achieve an accuracy close to the SVM accuracy is usually far less than the number of SVM support vectors.

Abstract translation: 支持向量机（SVM）虽然准确，但由于支持向量的数量庞大，因此在分类速度很高的应用中并不优选。 为了克服这个问题，已经设计了具有以下性质的原始系统和方法：（1）它将基函数的概念与支持向量的概念相分离; （2）它贪婪地找到一组指定的最大大小（d>最大）的内核基函数来很好地近似SVM原始成本函数; （3）它是有效的，并且粗略地作为O（nd max max）来缩放，其中n是训练样本的数量; 和（4）实现接近SVM精度的精度所需的基本函数的数量通常远小于SVM支持向量的数量。

公开(公告)号：US20060074908A1

公开(公告)日：2006-04-06

申请号：US10949821

申请日：2004-09-24

Applicant: Sathiya Selvaraj ,  Dennis DeCoste

Inventor： Sathiya Selvaraj ,  Dennis DeCoste

IPC: G06F17/30 ,  G06F15/18

CPC classification number: G06F17/30707 ,  G06K9/6269 ,  G06N7/005 ,  Y10S707/99931 ,  Y10S707/99936 ,  Y10S707/99937

Abstract: The present invention provides a system and method for building fast and efficient support vector classifiers for large data classification problems which is useful for classifying pages from the World Wide Web and other problems with sparse matrices and large numbers of documents. The method takes advantage of the least squares nature of such problems, employs exact line search in its iterative process and makes use of a conjugate gradient method appropriate to the problem. In one embodiment a support vector classifier useful for classifying a plurality of documents, including textual documents, is built by selecting a plurality of training documents, each training document having suitable numeric attributes which are associated with a training document vector, then initializing a classifier weight vector and a classifier intercept for a classifier boundary, the classifier boundary separating at least two document classes, then determining which training document vectors are suitable support vectors, and then re-computing the classifier weight vector and the classifier intercept for the classifier boundary using the suitable support vectors together with an iteratively reindexed least squares method and a conjugate gradient method with a stopping criterion.

Abstract translation: 本发明提供了一种用于构建用于大数据分类问题的快速和有效的支持向量分类器的系统和方法，其对于从万维网分类页面和用于稀疏矩阵和大量文档的其他问题是有用的。 该方法利用了这些问题的最小二乘性质，在其迭代过程中采用精确的线搜索，并利用了适合该问题的共轭梯度法。 在一个实施例中，通过选择多个训练文档来构建用于分类多个文档（包括文本文档）的支持向量分类器，每个训练文档具有与训练文档向量相关联的合适的数字属性，然后初始化分类器权重 向量和分类器边界的分类器截距，分类器边界分离至少两个文档类，然后确定哪些训练文档矢量是适合的支持向量，然后使用分类器边界重新计算分类器权重向量和分类器截距 合适的支持向量以及迭代重索引最小二乘法和具有停止标准的共轭梯度法。