Unlike descriptive statistics, where our goal is to describe various characteristics of a dataset using specific quantities, with inferential statistics, we'd like to perform a particular statistical modeling process on our dataset so that we can infer further information, either about the dataset itself or even about unseen data points that are from the same population.
In this section, we will go through a number of different methods of inferential statistics. From these discussions, we will see that each method is designed for specific data and situations, and it is the responsibility of the statistician or machine learning engineer to appropriately apply them.
The first method that we will discuss is one of the most fundamental in classical statistics: t-tests.
In general, t-tests (also known as Student's t-tests) are used to compare two mean (average) statistics and conclude whether they are different enough from each other...