Introduction to SVMs in Python for Data Science:
SVMs are supervised learning models which has associated learning algorithms that analyzes data and then use it for classification and regression analysis. SVM is a discriminative classifier which is formally defined by separating hyperplane. The given labeled training data, algorithm outputs an optimal hyperplane that categorizes new examples.
Support vector machines:
SVM model is a representation of examples as points in space, map so that examples of separate categories are divide by clear gap which is as wide as possible.
In order to perform linear classification, SVMs will efficiently perform non-linear classification, implicitly mapping their inputs into high dimensional feature spaces.
What SVM does?
SVM training algorithm will build a model which assigns new examples to one category or other. It is a non-probabilistic binary linear classifier.
In this, we will further discuss examples about SVM classification of cancer UCI datasets by use of machine learning tools, i.e., scikit-learn compatible with python.
Now we will observe an example of support vector classification. First, we will create a dataset.
SVM not only draws a line between two classes but also considers a region about the line of given width. Below is an example of how it will look like.
It is intuition of SVMs that optimize a linear discriminant model which represents the perpendicular distance between datasets. Now we will train the classifier using our training data. We need to import the cancer dataset as csv file before training. We will train two features out of all the features.
Fitting a Support Vector Machine:
Now we will fit a support vector machine classifier to these points. Here we will treat scikit-learn algorithm like a black box that will accomplish the task.
Now we will look at the graph.