【摘要】：為了有效地在半監(jiān)督多視圖情景下進(jìn)行維數(shù)約簡,提出了使用非負(fù)低秩圖進(jìn)行標(biāo)簽傳播的半監(jiān)督典型相關(guān)分析方法。非負(fù)低秩圖捕獲的全局線性近鄰可以利用直接鄰居和間接可達(dá)鄰居的信息維持全局簇結(jié)構(gòu),同時(shí)低秩的性質(zhì)可以保持圖的壓縮表示。當(dāng)無標(biāo)簽樣本通過標(biāo)簽傳播算法獲得估計(jì)的標(biāo)簽信息后,在每個(gè)視圖上構(gòu)建軟標(biāo)簽矩陣和概率類內(nèi)散度矩陣,然后通過最大化不同視圖同類樣本間相關(guān)性的同時(shí)最小化每個(gè)視圖低維特征空間類內(nèi)變化來提升特征鑒別能力。實(shí)驗(yàn)結(jié)果表明,所提方法比已有相關(guān)方法能夠取得更好的識(shí)別性能且更魯棒,是有效的多視圖維數(shù)約簡方法。
[Abstract]:In order to reduce dimensionality effectively in semi-supervised multi-view scenarios, a semi-supervised canonical correlation analysis method using non-negative low-rank graphs for label propagation is proposed. The global linear nearest neighbor captured by non-negative low rank graph can maintain the global cluster structure using the information of direct neighbor and indirect reachable neighbor, and the property of low rank can maintain the compressed representation of graph. When untagged samples obtain the estimated tag information through tag propagation algorithm, soft tag matrix and probabilistic intra-class divergence matrix are constructed on each view. Then the ability of feature identification is improved by maximizing the correlation between the same samples of different views and minimizing the intra-class changes in each view's low-dimensional feature space. The experimental results show that the proposed method can achieve better recognition performance and more robust than the existing methods, and it is an effective multi-view dimension reduction method.
【作者單位】： 九江學(xué)院信息科學(xué)與技術(shù)學(xué)院;南京郵電大學(xué)自動(dòng)化學(xué)院;
【基金】：國家自然科學(xué)基金資助項(xiàng)目(61462048) 九江學(xué)院科研項(xiàng)目(2014KJYB019,2014KJYB030,2015LGYB26) 江西省教育廳科學(xué)技術(shù)研究項(xiàng)目(GJJ151076)
【分類號(hào)】：TP391.41
，

本文編號(hào)：2159892