本文選題：小樣本集分類器　切入點(diǎn)：遷移學(xué)習(xí)　出處：《自動化學(xué)報》2016年09期

【摘要】：如何利用大量已有的同構(gòu)標(biāo)記數(shù)據(jù)(源域)設(shè)計小樣本訓(xùn)練數(shù)據(jù)(目標(biāo)域)的分類器是一個具有很強(qiáng)應(yīng)用意義的研究問題.由于不同域的數(shù)據(jù)特征分布有差異,直接使用源域數(shù)據(jù)對目標(biāo)域樣本進(jìn)行分類的效果并不理想.針對上述問題,本文提出了一種基于遷移學(xué)習(xí)的分類器設(shè)計算法.首先,本文利用內(nèi)積度量的邊際Fisher準(zhǔn)則對源域進(jìn)行特征映射,提高源域中類內(nèi)緊湊性和類間區(qū)分性.其次,為了篩選合理的訓(xùn)練樣本對,本文提出一種去除邊界奇異點(diǎn)的算法來選擇源域密集區(qū)域樣本點(diǎn),與目標(biāo)域中的標(biāo)記樣本點(diǎn)組成訓(xùn)練樣本對.在核化空間上,本文學(xué)習(xí)了目標(biāo)域特征到源域特征的非線性轉(zhuǎn)換,將目標(biāo)域映射到源域.最后,利用鄰近算法(k-nearest neighbor,k NN)分類器對映射后的目標(biāo)域樣本進(jìn)行分類.本文不僅改進(jìn)了邊際Fisher準(zhǔn)則方法,并且將基于自適應(yīng)樣本對篩選的遷移學(xué)習(xí)應(yīng)用到小樣本數(shù)據(jù)的分類器設(shè)計中,提高域間適應(yīng)性.在通用數(shù)據(jù)集上的實(shí)驗結(jié)果表明,本文提出的方法能夠有效提高小樣本訓(xùn)練域的分類器性能.
[Abstract]:How to use a large number of existing isomorphic tagged data (source domain) to design classifiers for small sample training data (target domain) is a significant research problem. The direct use of source domain data to classify target domain samples is not satisfactory. In view of the above problems, this paper proposes a classifier design algorithm based on migration learning. In this paper, the marginal Fisher criterion of inner product metric is used to map the source domain to improve the intra-class compactness and inter-class distinction in the source domain. Secondly, in order to select reasonable training sample pairs, In this paper, an algorithm to remove boundary singularity points is proposed to select the sample points in the dense region of the source domain and to form a training sample pair with the labeled sample points in the target domain. In the kernel space, we study the nonlinear transformation from the target domain features to the source domain features. The target domain is mapped to the source domain. Finally, we use the nearest nearest neighbor classifier to classify the target domain samples after mapping. This paper not only improves the marginal Fisher criterion method, The migration learning based on adaptive sample selection is applied to the classifier design of small sample data to improve inter-domain adaptability. The experimental results on the general data set show that, The proposed method can effectively improve the performance of small sample training domain classifier.
【作者單位】： 浙江大學(xué)信息與電子工程學(xué)院;浙江大學(xué)CAD&CG國家重點(diǎn)實(shí)驗室;
【基金】：國家自然科學(xué)基金(61471321) 教育部 中國移動科研基金(MCM20150503) 國家自然科學(xué)基金(61202400) 浙江省自然科學(xué)基金(LQ12F02014)資助~~
【分類號】：TP181

【相似文獻(xiàn)】

相關(guān)期刊論文 前4條

1 程劍;應(yīng)自爐;;一種基于Fisher準(zhǔn)則的二維主元分析表情識別方法[J];五邑大學(xué)學(xué)報(自然科學(xué)版);2006年02期

2 張瑞秋;張憲民;陳忠;;基于三維X射線和Fisher準(zhǔn)則的BGA焊點(diǎn)檢測算法[J];焊接學(xué)報;2011年04期

3 高全華;張士勇;孫鋒利;;基于Fisher準(zhǔn)則特征融合策略的PNN車牌漢字識別[J];現(xiàn)代電子技術(shù);2010年08期

4 ;[J];;年期

，

本文編號：1687391