ML concepts in brief: Supervised Learning

Supervised learning problems are generally of 2 types:

1. Classification problems: Target is to classify the data or label the data accurately after learning from a training data set.

Example of a Classification problem Dataset. After learning the features of a particular class, the goal of ML algorithm is to predict the class of a given random case with features.

2. Regression Problems: Target is to predict a numerical value/changing value after learning from a training dataset.

Example of a Regression problem Dataset. After learning the features of salary of all the professionals, the goal of ML algorithm is to predict the salary of a given random professional with random features.

How to decide whether a particular Machine Learning problem is a supervised learning problem:

Data for for both of the above problems is labelled i.e. it either has a target label (class) or a target value (salary) and our goal is to give out the same values (class/salary) for different test subjects after learning from the features of available dataset. THIS IS HOW WE DECIDE THAT A PROBLEM IS A SUPERVISED LEARNING PROBLEM.

Remove the label(class/salary column entirely) and the problem is converted to an unsupervised learning problem.