21120 - AI FOR MARKETING: MACHINE LEARNING AND CAUSAL INFERENCE
Department of Marketing
Course taught in English
Go to class group/s: 31
ACME (6 credits - II sem. - OP | SECS-P/08) - AFM (6 credits - II sem. - OP | SECS-P/08) - AI (6 credits - II sem. - OP | SECS-P/08) - CLMG (6 credits - II sem. - OP | SECS-P/08) - DSBA (6 credits - II sem. - OP | SECS-P/08) - EMIT (6 credits - II sem. - OP | SECS-P/08) - ESS (6 credits - II sem. - OP | ECON-07/A | SECS-P/08) - FIN (6 credits - II sem. - OP | SECS-P/08) - GIO (6 credits - II sem. - OP | SECS-P/08) - IM (6 credits - II sem. - OP | SECS-P/08) - MM (6 credits - II sem. - OP | SECS-P/08) - PPA (6 credits - II sem. - OP | SECS-P/08)
Course Director:
MARTON VARGA
MARTON VARGA
Mission & Content Summary
MISSION
Understanding concepts such as utility maximization, discrete choice, and consumer decision simulation is highly beneficial for aspiring managers and business analysts. This course equips students to predict and analyze various outcomes from different managerial actions. The course introduces modern machine learning methods for data-driven decision-making, enabling students to create artificial datasets that mirror real-world industry scenarios. By mastering advanced techniques and causal inference models, participants will analyze consumer behaviors, such as purchase patterns, subscription habits, churn rates, or sensitivity to targeted discounts. Guided by experienced data scientist mentors, students receive valuable feedback and insights throughout their group projects, refining their research questions and modeling approaches with managerial significance. This blend of theory and practice equips students to help firms make strategic, data-driven decisions, positioning them for success as future managers.
CONTENT SUMMARY
The course covers the foundations and applications of AI, machine learning, and causal inference for marketing decisions. Topics include data preparation and preprocessing, regression analysis, hypothesis testing, probabilities, and discrete choice models for understanding consumer behavior and managerial decision-making. Students will learn predictive machine learning methods such as cross-validation, variable selection, LASSO, regression trees, random forests, and classification techniques. The course also introduces causal inference and treatment effect analysis to evaluate the impact of managerial actions and marketing interventions. Throughout the course, students will apply these methods using R and work with simulated and real-world business data to develop data-driven insights for marketing and management contexts.
Intended Learning Outcomes (ILO)
KNOWLEDGE AND UNDERSTANDING
At the end of the course student will be able to...
Students will understand the main concepts and methods of machine learning, predictive analytics, and causal inference in marketing and business contexts. They will explain the differences between correlation, prediction, and causation, interpret the outputs of regression and classification models, and recognize the strengths and limitations of different analytical approaches. Students will also understand the role of data preprocessing, model validation, and simulation techniques in data analysis using R.
APPLYING KNOWLEDGE AND UNDERSTANDING
At the end of the course student will be able to...
Students will apply machine learning and causal inference methods to analyze managerial and marketing problems using R. They will evaluate predictive models, simulate consumer behavior under different business scenarios, and develop data-driven recommendations for managerial decisions. Students will also communicate analytical findings effectively, formulate relevant managerial research questions, and collaborate in teams to design and present analytical projects.
Teaching methods
- Lectures
- Guest speaker's talks (in class or in distance)
- Practical Exercises
- Collaborative Works / Assignments
DETAILS
|
Guest speakers Senior data scientist(s) with 10 years of industry experience will share their insights on working in a data science team, the importance of managerial acumen, and how to lead data teams efficiently.
Exercises
We will cover a variety of coding exercises in R during the classes.
Group assignments
Students will collaborate in groups on a project that involves creating a computer simulation of a business scenario and testing the outcomes of various managerial decisions. An experienced data scientist will supervise the project, providing feedback on methodology and managerial insights. This approach offers an excellent opportunity to learn about industry requirements, the relevance of data analytics, and the use of effective data inputs.
Interactive class activities For the group project, dedicated sessions will help student teams learn how to simulate data, form interesting managerial questions, and analyze the business scenario using predictions and data-driven simulations. |
Assessment methods
| Continuous assessment | Partial exams | General exam | |
|---|---|---|---|
|
x | ||
|
x | ||
|
x | ||
|
x |
ATTENDING STUDENTS
|
The final written exam will be based on the topics covered during class. It will evaluate students’ understanding and ability to apply the methods, models, and tools we have discussed. The exam does not include any elements related to computer coding, nor does it contain questions pertaining to coding. (Max 20 points)
Group Assignment:
Students will engage in a group project where their collaborative efforts will be assessed. The quality of group work, including the application of analytical methods, data interpretation, and presentation skills, will contribute to the overall evaluation. (Max 11 points)
Final Presentations Bonus Points:
A bonus point will be given to each member of the student group that delivers the best final presentation, as voted by their peers.
|
NOT ATTENDING STUDENTS
|
Not attending students will be evaluated through a written individual final exam, worth a maximum of 31 points. A minimum score of 18 points is required to pass the course. The exam is based on the required textbook indicated in the instructions for non-attending students. The exam does not include elements related to computer coding, nor does it contain questions requiring students to code.
The written individual exam verifies that students have achieved the expected learning outcomes by assessing their individual knowledge, skills, and abilities in applying machine learning and causal inference methods to business and marketing-related questions. In particular, students are expected to demonstrate their understanding of data analysis, regression, prediction and model selection, machine learning methods, treatment analysis, causal inference, difference-in-differences, and panel methods. They must be able to distinguish correlation from causation, interpret empirical results correctly, understand key identification assumptions, recognize potential threats to validity, compare alternative empirical approaches, and formulate appropriate empirical strategies for managerial questions.
The exam consists of open and/or closed answer questions. Closed-answer questions assess knowledge and comprehension of the main concepts, methods, assumptions, and interpretations covered in the course. Open-answer questions assess students’ ability to reason critically, apply the relevant methods to empirical and managerial problems, interpret results, and derive meaningful implications from data analysis. Conceptual clarity and reasoning are essential; memorization alone is not sufficient.
|
Teaching materials
ATTENDING STUDENTS
The material we cover during classes.
NOT ATTENDING STUDENTS
Chapters from: Békés, Gábor, and Gábor Kézdi. Data Analysis for Business, Economics, and Policy. Cambridge University Press, 2021. (the digital version of the book is available for free in the library)
Last change 22/05/2026 13:48