File:Hinge loss vs squared loss.png
Size of this preview: 560 × 599 pixels. Other resolution: 570 × 610 pixels.
Original file (570 × 610 pixels, file size: 25 KB, MIME type: image/png)
Summary
Description | English: The hinge loss function is zero on one side, where points are correctly classified, but rises linearly on the other, where they are incorrectly classified. This yields hard-margin SVM behavior when a perfect classifier is possible, and soft-margin behavior in the imperfect case. The squared (L2) loss function penalizes points classified both correctly and incorrectly, and prevents points from being ignored (their contribution to the error is rarely zero). |
Date | 5 February 2018( ) |
File source | Own Work |
Author | Alistair Wick |
Licensing
|
File history
Click on a date/time to view the file as it appeared at that time.
Date/Time | Thumbnail | Dimensions | User | Comment | |
---|---|---|---|---|---|
current | 23:41, 5 February 2018 | 570 × 610 (25 KB) | AlistairWick (talk | contribs) | Corrected and clarified loss functions, and added 0-1 loss | |
21:50, 5 February 2018 | 697 × 617 (24 KB) | AlistairWick (talk | contribs) | User created page with UploadWizard |
You cannot overwrite this file.
File usage
The following page uses this file: