基于边界鉴别分析的递归维数约简算法 - 控制与决策.pdf

绗?5 鍗 绗? 鏈 鎺 鍒 涓 鍐 绛 2010 骞 7 鏈 Vol. 25 No. 7 Control and Decision Jul. 2010 鏂囩珷缂栧彿: 1001-0920 (2010) 07-1088-05 鍩轰簬杈圭晫閴村埆鍒嗘瀽鐨勯€掑綊缁存暟绾︾畝绠楁硶 1 1 1 2 浠讳笘閿 , 鍚曚繆鎬€, 鑳℃檽濠 , 寮犳檽鍏 (1. 寰愬窞甯堣寖澶у璁＄畻鏈哄闄紝姹熻嫃寰愬窞221116 锛?. 涓浗鐭夸笟澶у鏈虹數宸ョ▼瀛﹂櫌锛屾睙鑻忓緪宸?21008) 鎽 瑕? 鎻愬嚭涓€绉嶅熀浜庤竟鐣岄壌鍒垎鏋愮殑閫掑綊缁存暟绾︾畝绠楁硶. 璇ョ畻娉曟妸宸叉眰鍙栬竟鐣岄壌鍒悜閲忔浜や簬寰呮眰瓒呭钩闈㈡硶鍚戦噺 浣滀负鏀寔鍚戦噺鏈?SVM) 浼樺寲闂鏂扮殑绾︽潫鏉′欢; 鐒跺悗瀵规敼杩汼VM 杩涜閫掑綊姹傝В, 寰楀埌姝ｄ氦杈圭晫閴村埆鍚戦噺鍩果 鏈€鍚 灏嗘暟鎹牱鏈湪姝ｄ氦杈圭晫閴村埆鍚戦噺涓婃姇褰卞疄鐜扮淮鏁扮害绠€. 璇ョ畻娉曚笉浠呭厠鏈嶄簡鐜版湁缁存暟绾︾畝绠楁硶闅句互鏀寔灏忔牱鏈暟鎹 闆嗐€佸彈鏁版嵁鏍锋湰鍒嗗竷褰卞搷绛夐棶棰? 鑰屼笖鎶藉彇鐨勭壒寰佸悜閲忓叿鏈夋洿浼樼殑鍒嗙被鎬ц兘. 浠跨湡瀹為獙璇存槑浜嗙畻娉曠殑鏈夋晥鎬? 鍏抽敭璇岤 鏀寔鍚戦噺鏈猴紱鍒嗙被锛涚淮鏁扮害绠€锛涜竟鐣岄壌鍒垎鏋 涓浘鍒嗙被鍙耳 TP391.14 鏂囩尞鏍囪瘑鐮? A Recursive dimension reduction algorithm based on margin discriminant analysis 1 1 1 2 REN Shi-jin , LV Jun-huai , HU Xiao-ting , ZHANG Xiao-guang (1. School of Computer锛孹uzhou Normal University 锛孹uzhou 221116 锛孋hina 锛?. School of Mechanical and Electrical Engineering 锛孋hina University Mining Technology锛孹uzhou 221008 锛孋hina. Correspondent 锛歊EN Shi-jin 锛孍-mail: sjren phd@163.com) Abstract: A recursive dimension reduction algorithm based on marginal discriminant analysis is presented. A constraint condition that the obtained margin discriminant vectors are orthogonal to the new normal vector of classi铿乧ation hyperplane is added to the optimal problem of support vector machines (SVM), and margin discriminant vectors can be recursively achieved by solving the modi铿乪d SVM. The number of dimension of data can be reduced by projecting data in orthogonal margin discriminant basis. The algorithm can not onl