When Riya’s elder sister plotted study hours against marks, she unwittingly performed linear regression. The table shows a clean 1‑to‑1 rise: 1 hour yields 20 marks, 2 hours 40, 3 hours 60. The pattern is unmistakable—input grows, output follows in a steady, predictable way.

Linear regression boils down to drawing a single straight line that best fits the dots of real data. Coefficients assign weight to each input—experience might carry 5,000, skills 3,000—while the intercept anchors the line at a base value. The model explains why numbers shift.

Because it is easy to understand, quick to train, and transparent to managers, linear regression dominates interview questions. Recruiters test whether candidates grasp the relationship, not just recite formulas. The method’s simplicity also makes it a first‑step tool in many data‑science pipelines.

The next chapter tackles model quality: how to judge a line’s fit and quantify its mistakes. Errors and loss functions reveal whether predictions drift from reality. In Day 3, the author promises a deeper dive into measuring model accuracy and refining the straight‑line hypothesis.