發(fā)布時(shí)間：2018-06-18 22:59

  本文選題：貝葉斯網(wǎng)絡(luò) + 混合結(jié)構(gòu)學(xué)習(xí)��； 參考：《山西財(cái)經(jīng)大學(xué)》2017年碩士論文

【摘要】：近年來(lái),貝葉斯網(wǎng)絡(luò)(Bayesian network,BN)在不確定性知識(shí)表示與概率推理方面發(fā)揮著越來(lái)越重要的作用。作為一種重要的圖模型方法,該模型已被廣泛應(yīng)用于生物信息學(xué)、金融預(yù)測(cè)分析、編碼學(xué)、數(shù)據(jù)挖掘與機(jī)器學(xué)習(xí)等領(lǐng)域。其中,BN結(jié)構(gòu)學(xué)習(xí)是BN推理研究中的重要問題,也是該模型推向應(yīng)用的前提基礎(chǔ)。然而,當(dāng)前較為流行的兩階段混合結(jié)構(gòu)學(xué)習(xí)算法中,大多存在兩個(gè)問題:第一階段無(wú)向超結(jié)構(gòu)學(xué)習(xí)中存在容易丟失弱關(guān)系的邊的問題;第二階段的爬山隨機(jī)搜索時(shí)存在易陷入局部極值的問題。針對(duì)這些問題,本文基于超結(jié)構(gòu)和隨機(jī)搜索策略,研究了兩種BN結(jié)構(gòu)的混合學(xué)習(xí)算法。具體研究?jī)?nèi)容和創(chuàng)新之處包括:第一,提出了基于超結(jié)構(gòu)和隨機(jī)搜索的SSRS算法。首先采用Opt01ss算法學(xué)習(xí)超結(jié)構(gòu),盡可能地避免出現(xiàn)丟邊現(xiàn)象。然后,給出基于超結(jié)構(gòu)的兩種隨機(jī)搜索操作,分析初始網(wǎng)絡(luò)的隨機(jī)產(chǎn)生規(guī)則和對(duì)初始網(wǎng)絡(luò)的隨機(jī)優(yōu)化策略,重點(diǎn)提出SSRS結(jié)構(gòu)學(xué)習(xí)算法,該算法在一定程度上可以很好地跳出局部最優(yōu)極值。第二,提出了擴(kuò)展的SSRS算法,即E-SSRS算法。在E-SSRS算法中,首先在初始網(wǎng)絡(luò)的選擇階段,增加了通過(guò)評(píng)分選擇對(duì)每條邊添加方向之步驟,使得選取的初始網(wǎng)絡(luò)更靠近最優(yōu)網(wǎng)絡(luò)。然后,在優(yōu)化階段,對(duì)刪邊策略進(jìn)行了擴(kuò)展,使用了基于馬爾科夫毯的策略對(duì)網(wǎng)絡(luò)進(jìn)行修剪,進(jìn)一步提出E-SSRS算法。通過(guò)擴(kuò)展,使該算法減少了搜索次數(shù),進(jìn)一步提高算法效率。第三,設(shè)計(jì)并實(shí)現(xiàn)了SSRS算法和E-SSRS算法。分別在標(biāo)準(zhǔn)Survey,Asia和Sachs,Child網(wǎng)絡(luò)上,通過(guò)幾種評(píng)價(jià)指標(biāo),并與已有六種混合學(xué)習(xí)算法實(shí)驗(yàn)結(jié)果的對(duì)比分析,驗(yàn)證了本文所提出的兩種混合結(jié)構(gòu)學(xué)習(xí)算法的良好性能。
[Abstract]:In recent years, Bayesian Network (BN) plays an increasingly important role in uncertain knowledge representation and probabilistic reasoning. As an important graph model method, the model has been widely used in the fields of bioinformatics, financial prediction and analysis, coding, data mining and machine learning. The learning of BN structure is an important problem in the research of BN reasoning, and it is also the prerequisite for the application of the model. However, most of the popular two-stage hybrid structure learning algorithms have two problems: the first stage undirected superstructure learning has the problem of losing the edge of weak relation easily in the first stage undirected superstructure learning; In the second stage of the mountain climbing random search, the problem of local extremum is easy to fall into. In order to solve these problems, the hybrid learning algorithms of two BN structures are studied based on hyperstructure and random search strategy. The main contents and innovations are as follows: first, a new SSRS algorithm based on hyperstructure and random search is proposed. First, we use Opt01ss algorithm to learn the superstructure to avoid the phenomenon of edge loss as far as possible. Then, two kinds of random search operations based on superstructure are given, the random generation rules of initial network and the random optimization strategy of initial network are analyzed, and the SSRS structure learning algorithm is put forward. To some extent, the algorithm can get rid of the local optimal extremum. Secondly, an extended SSRS algorithm, E-SSRS algorithm, is proposed. In the E-SSRS algorithm, in the selection stage of the initial network, the step of adding direction to each edge is added to make the selected initial network closer to the optimal network. Then, in the optimization phase, the edge deletion strategy is extended and the network is pruned based on Markov blanket, and an E-SSRS algorithm is proposed. By extending the algorithm, the search times are reduced and the efficiency of the algorithm is further improved. Thirdly, SSRS algorithm and E-SSRS algorithm are designed and implemented. In the standard survey Asia and Sachschild networks, the performance of the two hybrid learning algorithms proposed in this paper is verified by comparing the experimental results with the experimental results of six hybrid learning algorithms.
【學(xué)位授予單位】：山西財(cái)經(jīng)大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP18

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 劉建偉;崔立鵬;劉澤宇;羅雄麟;;正則化稀疏模型[J];計(jì)算機(jī)學(xué)報(bào);2015年07期

2 劉建偉;崔立鵬;羅雄麟;;概率圖模型的稀疏化學(xué)習(xí)[J];計(jì)算機(jī)學(xué)報(bào);2016年08期

3 陶卿;高乾坤;姜紀(jì)遠(yuǎn);儲(chǔ)德軍;;稀疏學(xué)習(xí)優(yōu)化問題的求解綜述[J];軟件學(xué)報(bào);2013年11期

4 冀俊忠;張鴻勛;胡仁兵;劉椿年;;基于蟻群算法的貝葉斯網(wǎng)結(jié)構(gòu)學(xué)習(xí)[J];北京工業(yè)大學(xué)學(xué)報(bào);2011年06期

5 許麗佳;黃建國(guó);王厚軍;龍兵;;混合優(yōu)化的貝葉斯網(wǎng)絡(luò)結(jié)構(gòu)學(xué)習(xí)[J];計(jì)算機(jī)輔助設(shè)計(jì)與圖形學(xué)學(xué)報(bào);2009年05期

6 張少中;王秀坤;丁華;;基于模擬退火的貝葉斯網(wǎng)絡(luò)結(jié)構(gòu)學(xué)習(xí)算法[J];計(jì)算機(jī)科學(xué);2004年10期

相關(guān)博士學(xué)位論文 前2條

1 劉峰;貝葉斯網(wǎng)絡(luò)結(jié)構(gòu)學(xué)習(xí)算法研究[D];北京郵電大學(xué);2008年

2 張少中;基于貝葉斯網(wǎng)絡(luò)的知識(shí)發(fā)現(xiàn)與決策應(yīng)用研究[D];大連理工大學(xué);2003年

，

本文編號(hào)：2037154