Changeset 188

Timestamp:

Oct 20, 2004, 5:54:56 PM (18 years ago)

Author:

Peter

Message:

docs

File:

• trunk/src/SVM.h (modified) (1 diff)

trunk/src/SVM.h

@@ -72 +72 @@
     ///
     /// Training the SVM following Platt's SMO, with Keerti's
-    /// modifacation. However the complete kernel is stored in memory. The
-    /// reason for this is speed. When number of samples N is large this is
-    /// not possible since the memory cost for the kernel scales N^2. In that
-    /// case one should follow the SMO and calculate the kernel elements
-    /// sequentially
+    /// modifacation. However the complete kernel is stored in
+    /// memory. The reason for this is speed. When number of samples N
+    /// is large this is not possible since the memory cost for the
+    /// kernel scales N^2. In that case one should follow the SMO and
+    /// calculate the kernel elements sequentially. Minimizing \f$
+    /// \frac{1}{2}\sum
+    /// y_iy_j\alpha_i\alpha_j(K_{ij}+\frac{1}{C_i}\delta_{ij}) \f$,
+    /// which corresponds to minimizing \f$ \sum w_i^2+\sum
+    /// C_i\xi_i^2 \f$
     ///