So far, we've focused on describing data we have in hand using descriptive statistics and understanding the rules of chance through probability. But often, our goal in machine learning and data analysis isn't just to describe the data we collected; it's to make informed guesses or decisions about a larger group (the population) based on that limited data (the sample). This process of drawing conclusions about populations from samples is known as statistical inference.

This chapter introduces the fundamental concepts of statistical inference. You will learn about:

• The basic idea of estimating population characteristics (like the average) from sample data, known as point estimation.
• How to quantify the uncertainty around these estimates using confidence intervals.
• The core logic behind hypothesis testing: a formal way to make decisions about claims regarding a population, including formulating null and alternative hypotheses and interpreting p-values.
• How these inferential ideas connect to evaluating the significance of results in machine learning.

We will build upon the descriptive statistics and probability concepts from previous chapters to understand how to make generalizations beyond the immediate data.