Inference with big data is central to business today, where evidenced-based decisions are highly valued. Doing this is difficult because real world situations are often complex and fast-paced, and data can be simultaneously "big" and yet imperfect. In the real world, one has to analyze data for different types of decisions and situations and is rarely in the position of choosing his/her ideal data or setting. This means that the quality of the data and the method with which one can make inferences vary greatly across contexts. Moreover, handling big data, where one cannot visualize the entire dataset and visually identify problems, requires knowledge of advanced regression modeling and post-estimation techniques. Business leaders in such situations need to extract useful insights from data with advanced statistical modelling, and to communicate these insights in a non-technical and intuitive way so that others can understand.
This class addresses these needs by teaching advanced statistical modelling with big data, and practicing communicating these ideas at all levels of technicality (or non-technicality). It takes a practical view of statistics and data analysis with large datasets and provides students with a range of advanced state-of-the-art statistical and basic machine learning tools to address economic questions. Two unique cases were developed especially for this class. One of the cases is an "open-ended" project in which students will be required to apply the tools they learn to build a statistical model for analysis and then design business strategy based on the evidence. Each case reflects a highly complex real-world situations and evolve across lectures in stages so that students learn advanced analytical skills in a concrete context-relevant setting. This class will benefit students who want to think creatively about how to apply the results of rigorous data analysis to economic decisions.
This course is case-focused and most of the analysis will be conducted in groups in class. Using a fun and hands-on approach, students build on the foundational tools they obtained in the Business Analytics courses and learn advanced applications by working with big data projects in a lab-like setting in class. Students will analyze the data using STATA, interpret the results, assess their credibility and applicability to the economic questions which motivated the analysis, and present evidence-driven business decisions in class. There are no exams. Prerequisite: Business Analytics II (DECS 431) (see syllabus).
Management Science Major