In the lecture of SVM, for the duality, I do not understand why you said that "G(w, alpha) is convex in w and linear hence concave in alpha". Indeed, we have already seen that a linear function is convex so why did you say it's linear hence concave ?
Note that the function f(x) is called concave if -f(x) is convex. Negative of a linear function is also a linear function, that is convex as you already seen before in lab 6. Therefore linear function is concave.
Convexity and concavity of G(w, alpha)
Hello,
In the lecture of SVM, for the duality, I do not understand why you said that "G(w, alpha) is convex in w and linear hence concave in alpha". Indeed, we have already seen that a linear function is convex so why did you say it's linear hence concave ?
Hi,
Linear functions are both convex and concave.
Note that the function f(x) is called concave if -f(x) is convex. Negative of a linear function is also a linear function, that is convex as you already seen before in lab 6. Therefore linear function is concave.
1
Add comment