Technical Study Guide: The Mathematical Pipeline of Linear Regression

This study guide provides a rigorous exploration of the linear regression mathematical pipeline, moving from basic data components to the iterative optimisation of predictive models.

1. Introduction: The Linear Regression Objective

The fundamental objective of linear regression is to identify the optimal straight line that predicts a response variable based on a specific input. By modelling the relationship between variables, we transform raw observations into a predictive mathematical framework.

Grounding Example:

• Input (x): Study Hours
• Output (y): Marks

The goal is to determine exactly how many marks a student is likely to achieve based on the number of hours they have dedicated to study.

2. Foundational Components: Variables and Observations

2.1 Variables (x and y)

• Child Concept: x is the clue; y is the thing we are trying to guess.
• Intuition: To formulate a prediction, we must distinguish between the information we already possess (the predictor) and the outcome we wish to discover (the target).
• Mathematical Formulas:
  • x=input
  • y=output
• Aura Definition: Independent variables are explanatory features, while dependent variables represent the target response to be predicted.

2.2 Data Observations

• Child Concept: These are examples taken from real life.
• Intuition: A model cannot learn without evidence; we collect specific pairs of inputs and outputs to understand the underlying pattern of the data.
• Mathematical Formulas: Each observation is represented as an ordered pair (xi,yi), where i denotes the index of the observation.