In the "machine learning", the Support Vector Machine (SVM) is a supervised learning model with associated algorithms that analyze data and recognize patterns, used for classification and regression analysis in Business Intelligence. The SVM Basic takes a set of input data and predicts, for each given input, which of the two classes of output belongs, so it is a binary linear non-probabilistic classifier (just choose between 2 options). Given a set of training examples, each marked as belonging to one of two categories, a training algorithm builds a model that assigns new examples in a category or other. This type of model is widely used for the analysis of models in which a set of data has to be classified into only two categories, fraud or no-fraud, yes or no to credit, etc. An SVM model is a representation of the examples (database with which the estimation was made) as points in space, so that assign the examples of separate categories that are generally divided by a defined space, space that has to be as wide as it is possible. The new input data will be classified in the same space and to predict which category it belongs to.
The Support Vector Machine it is based on the fact that each new data can be classified within the corresponding category based on the learning of the analyzed data. In the example below, the objects belong to only one class either green or red. The separation line defines a boundary on the right side where all objects are green and on the left where all objects are red.
Advertisements
In this other example we can observe the two categories divided by a central axis and two clearance distances between them.
Advertisements
In real life it is very difficult to find models as clear as those in these images, but it is always possible to have an approximation.
Advertisements
