語系:
繁體中文
English
說明(常見問題)
回圖書館首頁
手機版館藏查詢
登入
回首頁
到查詢結果
[ subject:"Artificial Intelligence." ]
切換:
標籤
|
MARC模式
|
ISBD
On semi-supervised Kernel methods.
~
Sindhwani, Vikas.
FindBook
Google Book
Amazon
博客來
On semi-supervised Kernel methods.
紀錄類型:
書目-語言資料,印刷品 : Monograph/item
正題名/作者:
On semi-supervised Kernel methods./
作者:
Sindhwani, Vikas.
面頁冊數:
190 p.
附註:
Adviser: Partha Niyogi.
Contained By:
Dissertation Abstracts International68-10B.
標題:
Artificial Intelligence. -
電子資源:
http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3287085
ISBN:
9780549301769
On semi-supervised Kernel methods.
Sindhwani, Vikas.
On semi-supervised Kernel methods.
- 190 p.
Adviser: Partha Niyogi.
Thesis (Ph.D.)--The University of Chicago, 2007.
Semi-supervised learning is an emerging computational paradigm for learning from limited supervision by utilizing large amounts of inexpensive, unsupervised observations. Not only does this paradigm carry appeal as a model for natural learning, but it also has an increasing practical need in most if not all applications of machine learning---those where abundant amounts of data can be cheaply and automatically collected but manual labeling for the purposes of training learning algorithms is often slow, expensive, and error-prone.
ISBN: 9780549301769Subjects--Topical Terms:
769149
Artificial Intelligence.
On semi-supervised Kernel methods.
LDR
:03299nam 2200289 a 45
001
958962
005
20110704
008
110704s2007 ||||||||||||||||| ||eng d
020
$a
9780549301769
035
$a
(UMI)AAI3287085
035
$a
AAI3287085
040
$a
UMI
$c
UMI
100
1
$a
Sindhwani, Vikas.
$3
1282429
245
1 0
$a
On semi-supervised Kernel methods.
300
$a
190 p.
500
$a
Adviser: Partha Niyogi.
500
$a
Source: Dissertation Abstracts International, Volume: 68-10, Section: B, page: 6747.
502
$a
Thesis (Ph.D.)--The University of Chicago, 2007.
520
$a
Semi-supervised learning is an emerging computational paradigm for learning from limited supervision by utilizing large amounts of inexpensive, unsupervised observations. Not only does this paradigm carry appeal as a model for natural learning, but it also has an increasing practical need in most if not all applications of machine learning---those where abundant amounts of data can be cheaply and automatically collected but manual labeling for the purposes of training learning algorithms is often slow, expensive, and error-prone.
520
$a
In this thesis, we develop families of algorithms for semi-supervised inference. These algorithms are based on intuitions about the natural structure and geometry of probability distributions that underlie typical datasets for learning. The classical framework of Regularization in Reproducing Kernel Hilbert Spaces (which is the basis of state-of-the-art supervised algorithms such as SVMs) is extended in several ways to utilize unlabeled data.
520
$a
These extensions are embodied in the following contributions: (1) Manifold Regularization is based on the assumption that high-dimensional data truly resides on low-dimensional manifolds. Ambient globally-defined kernels are combined with the intrinsic Laplacian regularizer to develop new kernels which immediately turn standard supervised kernel methods into semi-supervised learners. An outstanding problem of out-of-sample extension in graph transductive methods is resolved in this framework. (2) Low-density Methods bias learning so that data clusters are protected from being cut by decision boundaries at the expense of turning regularization objectives into non-convex functionals. We analyze the nature of this non-convexity and propose deterministic annealing techniques to overcome local minima. (3) The Co-regularization framework is applicable in settings where data appears in multiple redundant representations. Learners in each representation are biased to maximize consensus in their predictions through multi-view regularizers. (4) We develop l1 regularization and greedy matching pursuit algorithms for sparse non-linear manifold regularization . (5) We develop specialized linear algorithms for very large sparse data matrices, and apply it for probing utility of unlabeled documents for text classification. (6) Empirical results on a variety of semi-supervised learning tasks suggest that these algorithms obtain state-of-the-art performance.
590
$a
School code: 0330.
650
4
$a
Artificial Intelligence.
$3
769149
690
$a
0800
710
2
$a
The University of Chicago.
$3
1017389
773
0
$t
Dissertation Abstracts International
$g
68-10B.
790
$a
0330
790
1 0
$a
Niyogi, Partha,
$e
advisor
791
$a
Ph.D.
792
$a
2007
856
4 0
$u
http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3287085
筆 0 讀者評論
館藏地:
全部
電子資源
出版年:
卷號:
館藏
1 筆 • 頁數 1 •
1
條碼號
典藏地名稱
館藏流通類別
資料類型
索書號
使用類型
借閱狀態
預約狀態
備註欄
附件
W9122427
電子資源
11.線上閱覽_V
電子書
EB W9122427
一般使用(Normal)
在架
0
1 筆 • 頁數 1 •
1
多媒體
評論
新增評論
分享你的心得
Export
取書館
處理中
...
變更密碼
登入