This repository was archived by the owner on May 13, 2023. It is now read-only.

Commit 2e65746

authored

Update svm_gradient.py

1 parent 780f6ba commit 2e65746Copy full SHA for 2e65746

File tree

-18

lines changed

-18

lines changed

Lines changed: 4 additions & 18 deletions

Original file line number	Diff line number	Diff line change
`@@ -18,19 +18,9 @@ def svm_gradient(w, b, x, y, C):`
`18`	`18`
`19`	`19`	`"""`
`20`	`20`
`21`		`- grad_w = 0 # gradient`
`22`		`- grad_b = 0 # loss`
`23`		`-`
`24`		`-`
`25`		`- #######################################################################`
`26`		`- # TODO: #`
`27`		`- # Compute the gradient for a particular choice of w and b. #`
`28`		`- # Compute the partial derivatives and set grad_w and grad_b to the #`
`29`		`- # partial derivatives of the cost w.r.t. both parameters #`
`30`		`- # #`
`31`		`- #######################################################################`
`32`		`-`
`33`		`- lambda_ = 1/C`
	`21`	`+ grad_w = 0`
	`22`	`+ grad_b = 0`
	`23`	`+ lambda_ = 1/C`
`34`	`24`	`N = x.size`
`35`	`25`	`normw = np.linalg.norm(w)`
`36`	`26`	`w_t = np.transpose(w)`
`@@ -48,8 +38,4 @@ def svm_gradient(w, b, x, y, C):`
`48`	`38`	`var2[preds] = 0`
`49`	`39`	`grad_b = C * var2.sum()`
`50`	`40`
`51`		`-`
`52`		`- #######################################################################`
`53`		`- # END OF YOUR CODE #`
`54`		`- #######################################################################`
`55`		`- return grad_w, grad_b`
	`41`	`+ return grad_w, grad_b`

Comments

(0)