Linear regression as a basic model

Linear regression is a natural starting point for talking about machine learning. Not because it is "simple", but because it already contains almost everything: the model as a function, parameters, error, optimization, and geometric interpretation. If you understand linear regression, most other models will be perceived as its complications.

In the following cases we will treat linear regression not as a toy teaching algorithm, but as an engineering tool.

• Case 1. Apartment valuation based on parameters
• Case 2. Predicting a developer’s task completion time
• Case 3. Predicting server resource consumption
• Case 4. Estimating the likely customer check
• Case 5. Predicting market salary