[Convex Optimization] Soft Marin SVM & Introduction of Kernel Tricks

• Soft Margin SVM that provides flexibility by allowing some classification error by introducing some slack variable for each data point
• prevents over-fitting and enables finding decision boundary for linearly non-separable data
• Kernel tricks are used to construct decision boundary for linearly non-separable data by transforming the data into higher dimension without explicitly computing it.

Summary Notes

• Soft Marin SVM & Introduction of Kernel Tricks