Скачать с ютуб Complete Maths behind SVM And Kernel Trick|Support Vector Machines #2| в хорошем качестве

Complete Maths behind SVM And Kernel Trick|Support Vector Machines #2|

Support Vector Machines(SVM) are one of the most powerful machine learning algorithms , and even manage to outperform neural networks in some cases. In this video , i am going to cover the complete mathematics behind Support Vector machine algorithm , exploring what makes them so powerful. Topics covered: 1)How do we find output in SVM ? 2)How is optimization done in SVM ? 3)How do we find margin width of decision boundary ? 4) What is the Kernel Trick ? When to use which kernel ? Connect with me on LinkedIn:   / nachiketa-hebbar-86186515b   If you any doubts , leave it in the comments and thanks for watching ! Time Stamps if you want to skip to certain portions of the video 00:15 : Introduction 00:50 : How much maths should you know? 1:47: Dot Product in SVM 2:40 : Maths behind finding Output in SVM Model 6:45 : What to optimize in SVM? Margin Width in Depth 10:20 : Finding what the Margin of SVM Depends upon 12:02 : Introducing Lagrange Multiplier 15:24 : How SVM outperforms Neural Networks in Some cases 15:56 : Power of Convex Loss function in SVM 17:40 : Kernel Trick 19:15 : How to Decide which Kernel to use when ?