Course Wiki

Linear Regression

The regression model ML Zoomcamp builds from scratch: predicting a number from features with a weighted sum plus bias.

Module 2 builds a car-price prediction model with linear regression. The prediction is a weighted sum of feature values plus a bias term, and training means adjusting the weights so predictions land close to the actual prices. The course implements the math with NumPy first, then repeats it with scikit-learn.

Around the model, the module teaches the surrounding craft: exploratory data analysis, train/validation/test splits, feature preparation (one-hot encoding categorical variables, handling missing values), feature scaling, and RMSE as the error metric. Regularization is introduced here to control weight magnitudes when features are redundant.

The same model reappears in Stock Markets Analytics Zoomcamp for understanding data relationships in financial time series.

Taught in

• Machine Learning Zoomcamp

  • Module 2: Machine Learning for Regression
    • Car price prediction project
    • Linear regression
    • Linear regression: vector form
    • Training linear regression: Normal equation
    • Baseline model for car price prediction project
    • Car price prediction project summary
  • Module 3: Machine Learning for Classification
    • Logistic regression
    • Explore more

• MLOps Zoomcamp

  • Module 1: Introduction
    • Homework ## Related concepts

• Regularization

• Logistic Regression

• Classification Metrics

• Time Series Decomposition

• CRISP-DM