基于Winslow泛函的生成模型
發(fā)布時間:2022-07-20 19:09
概率生成模型,也叫作生成模型,是在機器學習和概率統(tǒng)計問題中的一類具有極高實際應用價值的模型。它的應用十分廣泛,可以用來對不同種類的數(shù)據(jù)進行建模,比如圖像,聲音,文本數(shù)據(jù),同時它能夠通過多種方式融入強化學習,所以在數(shù)據(jù)預測,圖片處理,文本生成等領域有廣泛的作用。但是如何設計一種高效且有效的生成模型,也是非常具有挑戰(zhàn)性的。生成模型的關鍵步驟就是對目標分布進行參數(shù)化估計。為了在一定程度上簡化討論,在本文中我們將主要關注通過極小化交叉熵(KL)的原理工作的生成模型。生成模型的種類非常多,但是主要能分為兩類,一類是構(gòu)造一個顯式的密度分布。在這些顯式的密度模型中,密度是可以計算處理的,所以模型的更新也是相對直接的。比如變分自編碼器。另一類生成模型沒有顯式地表示數(shù)據(jù)所在空間上的概率分布,相反,該模型提供了某種方式來減少與這種概率分布的直接交互。通常是直接提取樣本的能力,比如使用馬爾科夫鏈來隨機變換現(xiàn)有樣本的方法,以便從同一分布中獲得另一個樣本。特別的,有一類特別的具有顯式密度函數(shù)的生成模型,是基于定義兩個不同空間之間的連續(xù)非線性變換來構(gòu)造的,稱為流模型。換句話說,這類模型從一個簡單的分布出發(fā),將其與...
【文章頁數(shù)】:71 頁
【學位級別】:碩士
【文章目錄】:
摘要
Abstract
Chapter 1 Introduction
1.1 Source, background, and significance
1.2 Research Status
1.2.1 Worldwide Research Status
1.2.2 Domestic Research Status
1.3 Thesis Outline
Chapter 2 Generative model
2.1 Preliminaries and Notations
2.2 Minimizing KL divergence through flow maps
2.3 Stein variational gradient descent
2.4 Summary
Chapter 3 Adaptive gird method
3.1 Grid distribution based on the equidistribution principle
3.2 Grid distribution based on the variational principle
3.3 Iterative grid redistribution
3.4 Summary
Chapter 4 Generative model driven by Winslow functional
4.1 The choice of monitor function
4.2 Generative model based on Winslow map
4.2.1 Theoretical results
4.2.2 Algorithm
4.2.3 Numerical example
4.3 Generative model based on mapping G
4.3.1 Theoretical results
4.3.2 Algorithm
4.3.3 Numerical example
4.4 Summary
Chapter 5 Generative model implementation by DNN
5.1 Preliminaries
5.1.1 The Deep Ritz Method
5.1.2 Generative adversarial network
5.2 Architecture of neural network
5.3 Numerical examples
5.4 Summary
Conclusions
結(jié)論
References
Appendix
Acknowledgements
本文編號:3664592
【文章頁數(shù)】:71 頁
【學位級別】:碩士
【文章目錄】:
摘要
Abstract
Chapter 1 Introduction
1.1 Source, background, and significance
1.2 Research Status
1.2.1 Worldwide Research Status
1.2.2 Domestic Research Status
1.3 Thesis Outline
Chapter 2 Generative model
2.1 Preliminaries and Notations
2.2 Minimizing KL divergence through flow maps
2.3 Stein variational gradient descent
2.4 Summary
Chapter 3 Adaptive gird method
3.1 Grid distribution based on the equidistribution principle
3.2 Grid distribution based on the variational principle
3.3 Iterative grid redistribution
3.4 Summary
Chapter 4 Generative model driven by Winslow functional
4.1 The choice of monitor function
4.2 Generative model based on Winslow map
4.2.1 Theoretical results
4.2.2 Algorithm
4.2.3 Numerical example
4.3 Generative model based on mapping G
4.3.1 Theoretical results
4.3.2 Algorithm
4.3.3 Numerical example
4.4 Summary
Chapter 5 Generative model implementation by DNN
5.1 Preliminaries
5.1.1 The Deep Ritz Method
5.1.2 Generative adversarial network
5.2 Architecture of neural network
5.3 Numerical examples
5.4 Summary
Conclusions
結(jié)論
References
Appendix
Acknowledgements
本文編號:3664592
本文鏈接:http://www.sikaile.net/kejilunwen/yysx/3664592.html
最近更新
教材專著