To understand how advertising affects sales, a natural approach is to predict expected sales as a function of advertising and other relevant factors. This is an example of regression, a powerful and widely used data-analysis method—and the central topic of this course. Students will learn how to apply regression tools to complex, real-world problems with the dual goals of understanding data and predicting future outcomes. The emphasis is on building statistical intuition, mastering fundamental concepts and developing practical implementation skills in a programming language (R or Python), rather than memorizing mathematical formulas. Real examples are used throughout to demonstrate how these techniques operate in practice.
Topics include:
• linear regression;
• multiple regression;
• model checking and selection;
• generalized linear models (e.g., logistic regression);
• time-series models and forecasting;
• causal inference;
• resampling methods, including the bootstrap and cross-validation.
We will also discuss recent developments in AI, focusing on those that relate to the core ideas of the course.
