東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Machine Learning Methods Based on Di...

Zhao, Chenchao.

FindBook

Google Book

Amazon

博客來

Machine Learning Methods Based on Diffusion Processes.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Machine Learning Methods Based on Diffusion Processes./
作者:	Zhao, Chenchao.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2018,
面頁冊數:	151 p.
附註:	Source: Dissertation Abstracts International, Volume: 80-05(E), Section: B.
Contained By:	Dissertation Abstracts International80-05B(E).
標題:	Theoretical physics. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13804566
ISBN:	9780438728868

Machine Learning Methods Based on Diffusion Processes.
Zhao, Chenchao.

Machine Learning Methods Based on Diffusion Processes. - Ann Arbor : ProQuest Dissertations & Theses, 2018 - 151 p.

Source: Dissertation Abstracts International, Volume: 80-05(E), Section: B.

Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2018.

This thesis presents three distinct machine learning algorithms based on the mathematical formalism and physical idea of diffusion processes.

ISBN: 9780438728868Subjects--Topical Terms:

2144760
Theoretical physics.

Machine Learning Methods Based on Diffusion Processes.
LDR:04260nmm a2200349 4500 001 2201409
005 20190429062349.5
008 201008s2018 ||||||||||||||||| ||eng d
020 $a 9780438728868
035 $a (MiAaPQ)AAI13804566
035 $a (MiAaPQ)101545
035 $a AAI13804566
040 $a MiAaPQ $c MiAaPQ
100 1 $a Zhao, Chenchao. $3 3428128
245 1 0 $a Machine Learning Methods Based on Diffusion Processes.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2018
300 $a 151 p.
500 $a Source: Dissertation Abstracts International, Volume: 80-05(E), Section: B.
500 $a Adviser: Sergei Maslov.
502 $a Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2018.
520 $a This thesis presents three distinct machine learning algorithms based on the mathematical formalism and physical idea of diffusion processes.
520 $a First, the idea of using heat diffusion on a hypersphere to measure similarity has been previously proposed and tested by computer scientists, demonstrating promising results based on a heuristic heat kernel obtained from the zeroth order parametrix expansion; however, how well this heuristic kernel agrees with the exact hyperspherical heat kernel remains unknown. This thesis presents a higher order parametrix expansion of the heat kernel on a unit hypersphere and discusses several problems associated with this expansion method. We then compare the heuristic kernel with an exact form of the heat kernel expressed in terms of a uniformly and absolutely convergent series in high-dimensional angular momentum eigenmodes. Being a natural measure of similarity between sample points dwelling on a hypersphere, the exact kernel often shows superior performance in kernel SVM classifications applied to text mining, tumor somatic mutation imputation, and stock market analysis. The improvement in classification accuracy compared with kernels based on Euclidean geometry may arise from ameliorating the curse of dimensionality on compact manifolds.
520 $a Second, the effective dissimilarity transformation (EDT) on empirical dissimilarity hyperspheres is proposed and studied using synthetic and gene expression data sets. Iterating the EDT turns a static data distribution into a dynamical process purely driven by the empirical data set geometry and adaptively ameliorates the curse of dimensionality, partly through changing the topology of a Euclidean feature space Rn into a compact hypersphere Sn. The EDT often improves the performance of hierarchical clustering via the automatic grouping of information emerging from global interactions of data points. The EDT is not restricted to hierarchical clustering, and other learning methods based on pairwise dissimilarity should also benefit from the many desirable properties of EDT.
520 $a Finally, quantum time evolution exhibits rich physics, attributable to the interplay between the density and phase of a wave function. However, unlike classical heat diffusion, the wave nature of quantum mechanics has not yet been extensively explored in modern data analysis. We propose that the Laplace transform of quantum transport (QT) can be used to construct an ensemble of maps from a given complex network to a circle S1, such that closely-related nodes on the network are grouped into sharply concentrated clusters on S1. The resulting QT clustering (QTC) algorithm is as powerful as the state-of-the-art spectral clustering in discerning complex geometric patterns and more robust when clusters show strong density variations or heterogeneity in size. The observed phenomenon of QTC can be interpreted as a collective behavior of the microscopic nodes that evolve as macroscopic cluster "orbitals" in an effective tight-binding model recapitulating the network.
520 $a In summary, the three machine learning methods are based on three distinct diffusion processes. The dynamic diffusion processes serve as a promising foundation for future development in machine learning methods.
590 $a School code: 0090.
650 4 $a Theoretical physics. $3 2144760
650 4 $a Biophysics. $3 518360
690 $a 0753
690 $a 0786
710 2 $a University of Illinois at Urbana-Champaign. $b Physics. $3 3170744
773 0 $t Dissertation Abstracts International $g 80-05B(E).
790 $a 0090
791 $a Ph.D.
792 $a 2018
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13804566