计算机科学与探索
計算機科學與探索
계산궤과학여탐색
JOURNAL OF FRONTIERS OF COMPUTER SCIENCE & TECHNOLOGY
2014年
10期
1246-1253
,共8页
王立鹏%费飞%接标%张道强
王立鵬%費飛%接標%張道彊
왕립붕%비비%접표%장도강
子图挖掘%特征选择%图核降维%脑网络分类
子圖挖掘%特徵選擇%圖覈降維%腦網絡分類
자도알굴%특정선택%도핵강유%뇌망락분류
subgraph mining%feature selection%graph-kernel-based dimensionality reduction%brain network classification
脑网络分类在脑科学研究和脑疾病诊断等领域引起了学者们的广泛关注。目前大多数有关脑网络分类的研究都是以单个脑区或成对脑区之间的相关性作为分类特征,其缺点是不能反映多个脑区之间的拓扑结构信息。为克服上述缺点,提出了一种基于子图选择和图核降维的脑网络分类方法。具体包括:(1)分别从正类训练样本组及负类训练样本组中提取多个频繁子图,进而利用基于频度差的子图选择算法选取最具判别性的子图集;(2)基于上述过程中得到的子图集,利用图核主成分分析(graph-kernel-based principal component analysis,GK-PCA)方法对经过子图选择后的图数据进行特征提取;(3)利用支持向量机(support vector machine, SVM)在特征提取后的数据上进行分类。在真实的轻度认知障碍(mild cognitive impairment,MCI)脑网络数据集上对该方法进行了验证,实验结果表明了该方法的有效性。
腦網絡分類在腦科學研究和腦疾病診斷等領域引起瞭學者們的廣汎關註。目前大多數有關腦網絡分類的研究都是以單箇腦區或成對腦區之間的相關性作為分類特徵,其缺點是不能反映多箇腦區之間的拓撲結構信息。為剋服上述缺點,提齣瞭一種基于子圖選擇和圖覈降維的腦網絡分類方法。具體包括:(1)分彆從正類訓練樣本組及負類訓練樣本組中提取多箇頻繁子圖,進而利用基于頻度差的子圖選擇算法選取最具判彆性的子圖集;(2)基于上述過程中得到的子圖集,利用圖覈主成分分析(graph-kernel-based principal component analysis,GK-PCA)方法對經過子圖選擇後的圖數據進行特徵提取;(3)利用支持嚮量機(support vector machine, SVM)在特徵提取後的數據上進行分類。在真實的輕度認知障礙(mild cognitive impairment,MCI)腦網絡數據集上對該方法進行瞭驗證,實驗結果錶明瞭該方法的有效性。
뇌망락분류재뇌과학연구화뇌질병진단등영역인기료학자문적엄범관주。목전대다수유관뇌망락분류적연구도시이단개뇌구혹성대뇌구지간적상관성작위분류특정,기결점시불능반영다개뇌구지간적탁복결구신식。위극복상술결점,제출료일충기우자도선택화도핵강유적뇌망락분류방법。구체포괄:(1)분별종정류훈련양본조급부류훈련양본조중제취다개빈번자도,진이이용기우빈도차적자도선택산법선취최구판별성적자도집;(2)기우상술과정중득도적자도집,이용도핵주성분분석(graph-kernel-based principal component analysis,GK-PCA)방법대경과자도선택후적도수거진행특정제취;(3)이용지지향량궤(support vector machine, SVM)재특정제취후적수거상진행분류。재진실적경도인지장애(mild cognitive impairment,MCI)뇌망락수거집상대해방법진행료험증,실험결과표명료해방법적유효성。
Brain network classification methods have attracted a lot of attentions in the fields including brain science and brain disease diagnosis. However, most of existing studies on brain network classification use brain regions or the correlation between paired brain regions as classification features, which cannot reflect the topology information among multiple brain regions. To overcome the problem, this paper proposes a novel brain network classification method with subgraph selection and graph-kernel-based dimensionality reduction. Firstly, this method mines two groups of frequent subgraphs from positive and negative classes respectively, and then selects the most discriminative sub-graphs using the subgraph selection algorithm based on their respective frequencies difference. Secondly, it uses the graph-kernel-based principal component analysis (GK-PCA) method to perform feature extraction on the graph datawith selected subgraphs. Finally, it adopts the support vector machine (SVM) to perform classification on the data with extracted features, validates the proposed method on real brain network dataset of mild cognitive impairment (MCI), and the experimental results show the efficacy of the proposed method.