O'Reilly logo

Building Probabilistic Graphical Models with Python by Kiran R Karkera

Stay ahead with the world's most comprehensive technology and business learning platform.

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, tutorials, and more.

Start Free Trial

No credit card required

Structured prediction

In typical applications of machine learning classifiers, the classifier predicts a class for the target variable, such as spam/non-spam to classify an e-mail. Usually, each instance (such as an individual e-mail) is independent of the next or previous e-mail.

However, there are several classes of applications where the target variable is related to its neighbors. Take the case of image segmentation where a typical problem is that in a picture of a cow on a pasture, we want to classify each pixel (or super pixel, which is a contiguous group of pixels as described by the image processing literature) as cow or grass. Each super pixel has a few neighbors, and if the super pixel Sp is at the stomach region of the cow and surrounded ...

With Safari, you learn the way you learn best. Get unlimited access to videos, live online training, learning paths, books, interactive tutorials, and more.

Start Free Trial

No credit card required