Curriculum

Autumn 2024

41100, Applied Regression Analysis - Panos Toulis

To understand how price affects sales one idea is to model average sales as a function of price through an available dataset. This is regression, a powerful and widely used data analysis technique, and also the topic of this course. Students will learn how to use regression to analyze a variety of complex real-world problems, with the aim of understanding data and making data-informed decisions. Focus is placed on understanding the fundamental concepts and implementation issues in a programming language (R, or alternative). Topics include: (i) linear regression; (ii) model checking and selection; (iii) generalized linear models (e.g. logistic regression); (iv) time series models and forecasting; and (v) causal inference. This course emphasizes the practical applications of regression, and considers many real-world datasets and problems. Although the course used to have a substantial programming component, this year we will explore the use of AI (such as ChatGPT) to aid with code generation.

37105, Data Science for Marketing Decision Making - Giovanni Compiani

Marketing decisions in the era of big data and artificial intelligence (AI) are based on a statistical analysis of large amounts of transaction and customer data. Using such an analysis we can predict the profitability or ROI of different marketing decisions, such as pricing, customer targeting, or digital advertising. The goal of this class is to introduce modern data-driven marketing techniques and train the students as data scientists who can analyze data and make marketing decisions using state-of-the-art tools employed in the industry. We will cover a wide range of topics, including demand modeling, the analysis of household-level data, customer relationship management (CRM), and digital marketing. The focus is on predicting the impact of marketing decisions, including pricing, advertising, and customer targeting, on customer profitability and the ROI from a customer interaction. The students will get immersed in a workflow that begins with the initial processing of the raw data and ends with the implementation of the marketing decision. First, we will learn how to manage and process large databases. The key tools used include some key packages in R that are designed for big data processing. Second, we will discuss and apply modern statistical tools building on regression analysis, including some key tools from the machine learning literature. Finally, we will learn how to implement key marketing decisions based on the statistical analysis of the data.

Winter 2025

41207, Causal Inference for Business Applications - Panos Toulis

In recent years, causal inference has become essential for data-driven decision making, as these methods can protect against biases in traditional statistical modeling techniques. In this course, students will learn how to use various methods to draw causal inferences through practical experience and real-world data examples in areas such as policy, marketing and operations. Topics covered will include randomized A/B experiments, difference-in-differences, instrumental variables, and modern machine learning/AI tools. 

41100, Applied Regression Analysis - Katja Smetanina

This course is about regression, a powerful and widely used data analysis technique which is used to understand how different random quantities relate to one another. Students will learn how to use regression to analyze a variety of complex real-world problems, with the aim of gaining insights from the data and also to potentially predict future events. Focus is placed on the understanding of fundamental concepts and its implementation in a programming language (R, or alternative). Real-world examples are used throughout the course to illustrate the application of techniques. Topics covered include: (i) short review of simple linear regression; (ii) multiple regression and model checking and diagnostics; (iii) generalized linear models (e.g. logistic regression); (iv) time series models and forecasting. We will also discuss the use of machine learning in the context of regression.

 
32210, Generative Thinking (new course in 2023/2024) - Sanjog Misra

Generative AI is a pivotal advancement in the realm of artificial intelligence that has the potential to transform various industries, from entertainment and advertising to healthcare and education. A thorough understanding of Generative technologies, potential uses cases and the implications are going to be necessary conditions of success. This new technology can automate content creation, optimize processes, and personalize experiences but more broadly it has the capacity to synthesize new data, simulate scenarios, generate ideas, and even design novel solutions to complex problems., making it a crucial technology for managers across all domains.This course provides a comprehensive overview of Generative AI, covering its foundational technologies, practical applications, and broader societal implications. It aims to equip students with a solid understanding of Generative AI and its relevance to various fields, as well as cultivate critical thinking on the ethical, legal, and societal challenges related to its deployment. Each class will have a lecture and discussion portion as well as some practicum/exercise/demo portion. In addition, there will be a course projects where you will work on the development of a Generative AI solution.

 
35126, Quantitative Portfolio Management - Ralph Koijen

This course develops a framework to use quantitative methods to build and analyze investment strategies. We will take advantage of recent innovations in AI models and extensively use models such as GPT-4 (and related tools). You will get an in-depth understanding and hands-on experience how these methods are incredibly useful in the asset management industry and how they can transform the industry in the future. We will how these models are useful not only when applied to text data, but also to data that are unique to finance such as portfolio holdings and flows. We will also use the AI models to develop code to analyze big data (such as stock prices and returns, firm fundamentals, text data, portfolio holdings and flows) and to then predict returns, measure risk, estimate the valuation of firms, and ultimately build investment strategies. The final project requires you to develop and pitch a new investment strategy using this framework as well. The course will use Python as a coding language, but no prior knowledge of Python is required for this course. The course starts from a brief review of the traditional portfolio choice framework introduced in the Investments course and then covers much of the recent research on quantitative methods to build and critically assess investment strategies.

30135, AI and Financial Information (new course) - Bradford Levy

This course will cover applications of artificial intelligence (AI) for analyzing financial information. A key feature of financial information is that it exists in a variety of formats: structured vs. unstructured, numerical vs. narrative, etc. As a result, financial information can be difficult to process using any single methodological approach. The applications covered will highlight the ability of AI to assist with analyzing financial information and focus on cutting-edge methods currently employed by top regulators, auditors, and hedge funds.

41201, Big Data - Veronika Rockova

BUS 41201 is a course about data mining: the analysis, exploration, and simplification of large high-dimensional datasets. Students will learn how to model and interpret complicated `Big Data' and become adept at building powerful models for prediction and classification. Techniques covered include an advanced overview of linear and logistic regression, model choice and false discovery rates, multinomial and binary regression, classification, decision trees, factor models, clustering, the bootstrap and cross-validation. We learn both basic underlying concepts and practical computational skills, including techniques for analysis of distributed data. Heavy emphasis is placed on analysis of actual datasets, and on development of application specific methodology. Among other examples, we will consider consumer database mining, internet and social media tracking, network analysis, and text mining.

 
37906, Applied Bayesian Econometrics - Sanjog Misra

This course will discuss applications of Bayesian methods to micro-econometric problems. We will particularly focus on issues pertaining to panel data models with unobserved heterogeneity and the use of hierarchical models to dealing with them. While the course is more generally useful, the applications and illustrations will be focused on Marketing and Industrial Organization.

Spring 2025

35137, Machine Learning in Finance (new course) - Leland Bybee
 
41204, Machine Learning - Christian Hansen and Damien Kozbur

This course aims to provide a high-level overview of machine learning and its applications in the business world. The goal of the course is to offer a broad, relatively non-technical introduction to key ideas from machine learning without digging too far into the mathematical minutiae. We hope the course will be accessible to students with an understanding of basic statistical concepts and mathematics but no deeper training. While one cannot do machine learning without coding, the course is not about coding and will not require any coding expertise. Overall, we hope students will leave the course able to have informed conversations about machine learning, understanding key conceptual ideas, and having a better idea of realistic opportunities for the use of machine learning in business as well as understanding potential challenges. We will cover the three basic learning paradigms: (i) supervised learning, (ii) unsupervised learning, and (iii) reinforcement learning. We will illustrate the paradigms through business relevant applications that may include detecting spam in email, click-through rate prediction in online advertisement, image classification, face recognition, sentiment analysis, churn prediction, algorithmic trading, customer segmentation, recommender systems, graph and time series mining, and anomaly detection. Through the applications, we will highlight important concepts and practical considerations. After covering basic learning methods, we will explore additional topics such as large language models, causal machine learning, and ethical considerations such as privacy and fairness.

 
41903, Applied Econometrics - Christian Hansen

In this Ph.D.-level course, we will discuss statistical methodology for use in causal inference in econometric settings. The course will cover topics such as basic treatment effect models (the linear model, linear IV model, linear panel data models), an introduction to nonparametric estimation, and estimation of treatment effects when effects may be heterogeneous. It is assumed that students are familiar with linear regression and methods of performing classical inference using asymptotic approximations for linear and nonlinear models.