What is a binary classifier and how to create it?

14 次查看(过去 30 天)
I have to create binary classifier based on multiple features, like Intensity, Position, etc. I am new to matlab, and dont know what exactly a binary classifier is? How can I create a binary classifier using two or more features?
I am not sure but, does it mean a matrix which values are 1 (true) for only those pixels which satisfy the given range of intensity and predefined given area in image?
Or here two binary classifiers will be created, one for intensty and one for position?
Please help me to understand the concept and how can I do this.
  2 个评论
Adam
Adam 2016-1-28
编辑:Adam 2016-1-28
In simplest terms a binary classifier is just something that classifies a group of objects into one of two defined sets according to some rule or other.
Depending on the specific binary classifier this could be simple or very complex.
e.g. a simple classifier would be heads/tails for tossing a coin. Every coin toss will result in one or the other
a more complex binary classifier might be one with a logical expression defining the sets based on combination of numerous criteria.
e.g. A group is defined by all objects for with a > 10 && b < 6 && z == true
for some numeric properties a and b and logical property z of each object.
How to create one depends entirely on what you want it to classify.
Nitinkumar Ambekar
Nitinkumar Ambekar 2016-1-28
Thank you @Adam, I am working on MRI image, in which I have to classify knee cartilage based on its Intensity, Position, Three jet, Eigenvalue, Eigenvector, as features. Just for now I want to try with only two features - Intensity and Position.
Would you please explain little bit more by considering the above case?

请先登录,再进行评论。

回答(1 个)

Image Analyst
Image Analyst 2016-1-28
It sounds kind of like you're talking about logistic regression. In statistics, logistic regression, or logit regression, or logit model is a regression model where the dependent variable is categorical. Binary dependent variables are where your variable can take only two values, such as pass/fail, win/lose, alive/dead, healthy/diseased, or cartilage/non-cartilage. You can have more than two categories though, like cartilage/bone/other tissue/air.
Let us know if that's the situation.

类别

Help CenterFile Exchange 中查找有关 Get Started with Statistics and Machine Learning Toolbox 的更多信息

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by