Detecting outliers using One-Class Support Vector Machine (OCSVM)

Support Vector Machine (SVM) is a popular supervised machine learning algorithm that is mainly known for classification but can also be used for regression. The popularity of SVM comes from the use of kernel functions (sometimes referred to as the kernel trick), such as linear, polynomial, Radius-Based Function (RBF), and the sigmoid function.

In addition to classification and regression, SVM can also be used for outlier detection in an unsupervised manner, similar to KNN, which is mostly known as a supervised machine learning technique but was used in an unsupervised manner for outlier detection, as seen in the Outlier detection using KNN recipe.

How to do it...

In this recipe, you will continue to work with the tx DataFrame, created in the Technical requirements section, to detect outliers using the ocsvm class from PyOD:

1. Start by loading the OCSVM class:

from pyod.models.ocsvm import OCSVM

1. There are a few parameters...