Supervised Learning

Supervised learning

Machine learning is gaining increasing importance in the field of business analytics and business intelligence today. Based on the type of data and business problem, machine learning approaches can be classified into supervised, unsupervised and semi-supervised learning. This module is designed to help understand how to apply machine learning on different types of data using SPSS and R software.

Classification in supervised learning

The primary difference between supervised, unsupervised and semi-supervised learning is the format of data. Supervised learning includes structured or formatted data. The prominent feature of supervised learning is that it inculcates the applications for classifiers. Classification helps to predict the categories and classifies data into categories based on a sample dataset.

Correlation in supervised learning

Supervised learning serves three main requisites, risk minimization, prediction of trends, and offering optimised e-solutions. Correlation is a statistical measure that helps in determining the extent of the relationship between two or more variables or factors. There are two types of correlations; bivariate and partial correlations. Some of the commonly used methods are:

Regressions in supervised learning

Extensive sets of data can be processed and predicted based on past trends and external information. Regression is applied to data in order to determine the relationship between a set of variables. In supervised machine learning, a regression can be performed using many tests, depending on the aim and type of data. Most popular regression tests are:

Detection in supervised learning

Machine learning is gaining increasing importance in the field of business analytics and intelligence today. It can be described as the ability of computers to learn without being explicitly programmed. Outlier detection and determination of values based on a training dataset is known as detection in supervised learning. It is used either as a replacement for regression or as a supplement. Common methods include:

Estimation in supervised learning

Supervised learning consists of algorithms generated through external information or instances to build hypothesis suiting the nature of the study. A critical part of supervised machine learning includes estimation of factors such as error rates, demand and revenues. Common methods of estimation include: