supportvectormachine支持向量機(jī)-資料下載頁
2024-10-17 21:59本頁面
【導(dǎo)讀】非線性最優(yōu)分類面SVM. SVM是一種基于統(tǒng)計(jì)學(xué)習(xí)理論的模式識(shí)別。來，現(xiàn)在已經(jīng)在許多領(lǐng)域（生物信息學(xué)，分開的兩類數(shù)據(jù)點(diǎn)距離分類面最遠(yuǎn)。求解該問題，得到分類器。期望風(fēng)險(xiǎn)R要依賴聯(lián)合概率F(x,y)的信。息，實(shí)際問題中無法計(jì)算。需要找到經(jīng)驗(yàn)風(fēng)險(xiǎn)最小和推廣能力最大的。對(duì)分類面方程g=wx-b應(yīng)滿足?？瞻鬃畲螅ú蛔儯?。將上述問題表示成拉格朗日乘子式。轉(zhuǎn)化為線性可分。使用函數(shù),將所有樣本點(diǎn)映射到高
【正文】 ?????????00][21m a xNNNNTN D ??? ?? 21BNBTNNBNTB DD ???? ?常數(shù)項(xiàng) 工作集的大小可以人為指定 相等項(xiàng) SVM的分解算法 ? Proposition(Build down): moving a variable from B to N leaves the cost function unchanged,and the solution is feasible in the subproblem ? Proposition(Build up) moving a variable that violates the optimality condition from N to B gives a strict improvement in the cost function when the subproblem is reoptimized SVM的分解算法 1. Arbitrarily choose |B| points from the data set. 2. Solve the subproblem defined by the variable in B. 3. While there exist some ?j j?N,such that replace any ?i ,i ? B,with ?j and solve the new subproblem. 1)(01)(1)(0???????jjjjjjjjjyxgandCyxgandCyxgand???SVM分解算法的實(shí)例 ? SVMlight Thorsten Joachims (UniversityDortmund ,Informatik, AIUnit) Make LargeScale SVM Learning Practical ? SMO John C. Platt (Microsoft Research) Fast Training of Support Vector Machines using Sequential Minimal Optimization 謝謝
教學(xué)課件相關(guān)推薦
鄂ICP備17016276號(hào)-1