30514 - BIG DATA FOR BUSINESS ANALYTICS
Course offered to incoming exchange students
Department of Decision Sciences
Supported by Siemens
EMANUELE BORGONOVO
Mission & Content Summary
MISSION
CONTENT SUMMARY
- The Principles of Machine Learning.
- Formulation of quantitative models via Linear Programs.
- The symplex method, Duality.
- Sensitivity Analysis.
- Network Type Problems.
- Big Data and Lasso: The Dantzig Selector.
- Industry 4.0 and Descriptive Analytics: Business Case studies.
These case studies are discussed and solved in the presence of guest lecturers from several international leading companies.
Intended Learning Outcomes (ILO)
KNOWLEDGE AND UNDERSTANDING
- Formulate a computational model to solve business and management problems.
- Appreciate the principles and the solution algorithms of linear programs at the basis of dedicated software for their application.
- Distinguish the wide range of business problems whose solution is supported by computational models.
- Recognize the challenges posed to quantitative methods by large dimensionality and big data and identify the corresponding technological solutions.
APPLYING KNOWLEDGE AND UNDERSTANDING
- Organize information to build a quantitative model in line with the input posed.
- Translate a business problem into a corresponding computational modelling frame.
- Use dedicated software in order to obtain quantitative information.
- Interpret solutions derived from implementing the chosen model in order to make optimal decisions.
- Analyze quantitative models with sensitivity analysis tools to obtain "managerial insights".
Teaching methods
- Face-to-face lectures
- Online lectures
- Guest speaker's talks (in class or in distance)
- Case studies /Incidents (traditional, online)
DETAILS
The course makes use of a combination of teaching techniques. Remote (online) but synchronous lectures are used for the sessions in which methodological and theoretical parts of the paper are proposed and discussed.
- In these sessions students are assisted in identifying the quantitative model, in implementing the model through dedicated software and in performing sensitivity analysis.
- In the second part of the course, students are exposed to the solution of industry case studies presented in a triplet of lectures. After the exposition by the experts of the industrial problem, participants are introduced to the methods of solution and are guided in critically discussing the results, the methodologies adopted and in identifying weaknesses and remaining open questions.
Assessment methods
Continuous assessment | Partial exams | General exam | |
---|---|---|---|
|
x | ||
|
x |
ATTENDING AND NOT ATTENDING STUDENTS
The assessment will be written in the form of: Assignments plus Final Respondus test. Students can take the exam in the following form, mode A or mode B at their choice.
Mode A: There will be one two intermediate assignments concerning the first part of the course, one assignment concerning the second part of the course, and a final general exam through the Responds platform.
- The first two assignments concern the material explained first part of the course. They consist of a series of quantitative questions in closed form. Students solve mathematical problems divided into open-ended numerical questions and multiple-choice (or multiple-answer) questions. The first assignments wish to test student's ability to formulate a computational model to solve a given business problem and their understanding of the principles of the solution of linear programming problems.
- The third assignment concerns the material explained in the second part of the course. It consists of closed-form/multiple-choice questions concerning the methodological aspects of the case studies and the material developed in the second part of the course. Students will be tested on their knowledge of the business problems illustrated in the second part of the course and on their ability to recognize the challenges of the use of quantitative models to solve such problems.
In Mode B students take a unique final assignment on the entire course program. This assignment is, virtually, the union of the assignments in Mode A.
With the assignments, students will also be tested on the dedicated subroutines developed for the course and implemented in the software Matlab.
All students will then be required to take a final Respondus test, which will test their knowledge of the entire course program, through multiple-choice/multiple-answer questions. The Respondus exam is closed book.
Teaching materials
ATTENDING AND NOT ATTENDING STUDENTS
- R.J. VANDERBEI, Linear Programming, Springer Series in Operational Research and Management Science, 2014, Fourth Edition, ISBN 978-1-4614-7629-0.
- F.S. HILLIER and G.J. LIEBERMAN, Introduction to Operations Research, Second Edition, 2001.
- Notes and slides provided by the teachers.