【摘要】：在全球信息化大潮的推動下,數(shù)據(jù)量日益增長,其增速讓人驚嘆,面對如此之多的信息和數(shù)據(jù)突如其來的沖擊大腦,人們沒有時間看這些數(shù)據(jù),人們的關(guān)注已經(jīng)轉(zhuǎn)到可貴的應(yīng)付手段上。因此,人們必須找到有效方法,應(yīng)對處理大量的信息以此來高效、準(zhǔn)確的找到有價值的信息。首先,本文針對具有相關(guān)性的數(shù)據(jù),考慮數(shù)據(jù)之間具有層次關(guān)系,將與目標(biāo)領(lǐng)域相關(guān)的數(shù)據(jù)根據(jù)相關(guān)性按不同比例調(diào)整權(quán)重,在經(jīng)典遷移學(xué)習(xí)算法Tr Ada Boost中引入層次相關(guān)性、權(quán)重調(diào)整策略、錯分樣本權(quán)重約束,從而提出一種結(jié)合層次相關(guān)性的遷移學(xué)習(xí)算法,以解決具有層次關(guān)系的數(shù)據(jù),高效調(diào)整權(quán)重的問題。在實驗部分,本文將基于層次相關(guān)性的遷移學(xué)習(xí)算法與SVM、Tr Adaboost算法的正確率、查準(zhǔn)率、查全率進(jìn)行對比,對實驗數(shù)據(jù)分析得到基于層次相關(guān)性的數(shù)據(jù)具有更準(zhǔn)確的分類結(jié)果。針對目標(biāo)領(lǐng)域與多個源領(lǐng)域都相近,但源領(lǐng)域與目標(biāo)領(lǐng)域不一樣,多個源領(lǐng)域也都不一樣,都是一個大的源領(lǐng)域下面的多個小的源領(lǐng)域領(lǐng)域,這些小的源領(lǐng)域還有些相關(guān)性,從多個小的源領(lǐng)域往一個目標(biāo)領(lǐng)域遷移。針對具有多個源領(lǐng)域,單個源領(lǐng)域樣本不足,負(fù)遷移的問題,提出一種多源實例遷移學(xué)習(xí)算法,考慮多個源領(lǐng)域的知識,使目標(biāo)領(lǐng)域能綜合考慮運(yùn)用每個源領(lǐng)域的知識,該方法先對源領(lǐng)域和目標(biāo)領(lǐng)域做并集訓(xùn)練出分類器,經(jīng)過測試后保留那些取并集后提升分類效果的源領(lǐng)域。然后,再將留下的源領(lǐng)域做并集,做Tr Adaboost遷移學(xué)習(xí),測試后按照指定的規(guī)則選擇最終的集合作為源領(lǐng)域,與目標(biāo)領(lǐng)域共同訓(xùn)練出分類器。最后,通過實驗對第三章,第四章所提出的基于層次相關(guān)性的遷移學(xué)習(xí)和多源實例遷移學(xué)習(xí)進(jìn)行詳細(xì)的描述,并將實驗結(jié)果做出客觀的分析總結(jié),對比改進(jìn)前后的正確率,查準(zhǔn)率,查全率。證明了基于層次相關(guān)性的遷移學(xué)習(xí)算法和多源實例遷移學(xué)習(xí)算法能夠非常有效地提高分類器的分類效果。
[Abstract]:Driven by the tide of global information technology, the amount of data is growing day by day, and the growth rate is amazing. In the face of so much information and data suddenly coming to the brain, people do not have time to look at these data. Attention has shifted to valuable coping techniques. Therefore, people must find an effective way to deal with a large amount of information in order to find valuable information efficiently and accurately. First of all, considering the hierarchical relationship between the data, this paper adjusts the weight of the data related to the target domain according to the correlation according to different proportions, and introduces the hierarchical correlation into the classical migration learning algorithm Tr Ada Boost. In order to solve the problem of data with hierarchical relationship and adjust weights efficiently, a transfer learning algorithm combining hierarchical correlation is proposed. In the experiment part, we compare the transfer learning algorithm based on hierarchical correlation with the correct rate, precision rate and recall rate of SVM,Tr Adaboost algorithm, and get more accurate classification results by analyzing the experimental data. The target domain is similar to multiple source domains, but the source domain is not the same as the target domain, and many source domains are not the same, they are many small source fields under a large source domain, and these small source domains have some relevance. Migrate from multiple small source domains to one target domain. In order to solve the problem of multiple source domains with insufficient samples and negative migration, a multi-source instance transfer learning algorithm is proposed, in which the knowledge of multiple source domains is considered, so that the target domain can comprehensively consider the use of knowledge in each source domain. The method firstly combines the source domain and the target domain to train the classifier. After the test, the source domain that is used to improve the classification effect is reserved. Then, the source domain is merged and the Tr Adaboost migration learning is done. After testing, the final set is selected as the source domain according to the specified rules, and the classifier is trained together with the target domain. Finally, the transfer learning based on hierarchical correlation and multi-source instance transfer learning proposed in chapter 3 and chapter 4 are described in detail through experiments, and the experimental results are objectively analyzed and summarized, and the correct rate before and after the improvement is compared. Precision rate It is proved that the hierarchical correlation based migration learning algorithm and the multi-source instance migration learning algorithm can effectively improve the classification effect of the classifier.
【學(xué)位授予單位】：遼寧大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP181

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 王勇;;云計算環(huán)境下信息保護(hù)與防泄露系統(tǒng)設(shè)計及實現(xiàn)[J];中國管理信息化;2017年01期

2 潘炯光;韋余永;;一種基于領(lǐng)域語義相關(guān)性挖掘的遷移學(xué)習(xí)方法[J];西南師范大學(xué)學(xué)報(自然科學(xué)版);2016年05期

3 張博;史忠植;趙曉非;張建華;;一種基于跨領(lǐng)域典型相關(guān)性分析的遷移學(xué)習(xí)方法[J];計算機(jī)學(xué)報;2015年07期

4 黃宜華;;大數(shù)據(jù)機(jī)器學(xué)習(xí)系統(tǒng)研究進(jìn)展[J];大數(shù)據(jù);2015年01期

5 尹寶才;王文通;王立春;;深度學(xué)習(xí)研究綜述[J];北京工業(yè)大學(xué)學(xué)報;2015年01期

6 何清;李寧;羅文娟;史忠植;;大數(shù)據(jù)下的機(jī)器學(xué)習(xí)算法綜述[J];模式識別與人工智能;2014年04期

7 張倩;李海港;李明;程玉虎;;基于多源動態(tài)TrAdaBoost的實例遷移學(xué)習(xí)方法[J];中國礦業(yè)大學(xué)學(xué)報;2014年04期

8 吳肖炎;郭瑞;;基于機(jī)器學(xué)習(xí)的計算機(jī)安全技術(shù)綜述(上)[J];保密科學(xué)技術(shù);2013年03期

9 ;大數(shù)據(jù)時代[J];中國電子科學(xué)研究院學(xué)報;2013年01期

10 賀英杰;葉宗民;金吉學(xué);;機(jī)器學(xué)習(xí)在入侵檢測中的應(yīng)用綜述[J];計算機(jī)安全;2010年03期

相關(guān)碩士學(xué)位論文 前6條

1 嚴(yán)海銳;多源遷移學(xué)習(xí)算法研究[D];華南理工大學(xué);2016年

2 王欣;基于遷移學(xué)習(xí)的跨領(lǐng)域推薦的方法研究[D];杭州電子科技大學(xué);2015年

3 韋余永;基于實例與特征的遷移學(xué)習(xí)文本分類方法研究[D];西南大學(xué);2015年

4 郭世碩;多源遷移學(xué)習(xí)的研究[D];西安電子科技大學(xué);2014年

5 沈鍵;電子商務(wù)的個性化協(xié)同過濾推薦算法研究[D];上海交通大學(xué);2013年

6 戴文淵;基于實例和特征的遷移學(xué)習(xí)算法研究[D];上海交通大學(xué);2009年

，

本文編號：2394269