Course 2024-2025 a.y.

• Decision analysis: structuring a business problem with models and data.
• Principles of Machine Learning: the structure of a Data Science Problem
• Predictive models for a continuous response: linear regression.
• Predictive models for a categorical response: logistic regression.
• Explorative Methods for Multivariate Data: data exploration, principal component analysis and clustering
• Machine Learning Approaches: Ridge and Lasso Regression, Decision Trees and Random Forests

• Recognize appropriate models to structure business and decision problems.
• Identify the correct methodology for solving business and management problems through data science.
• Discern between alternative machine tasks and alternative models.

• Organize information to build a quantitative model in line with the input posed.
• Translate a decision problem into a corresponding quantitative model.
• Use the software R and Silver Decisions in order to determine solutions to a problem.
• Interpret solutions derived from implementing the chosen model in order to obtain managerial insights.
• Analyze models with tools that come from the statistical theory to make inference robust.

• Lectures

Teaching and learning activities for this course are divided into face-to-face lectures during which management problems are explained and solution models through quantitative methods are proposed and discussed. Students are assisted in:

• Identifying the decision analysis, statistical or machine learning model, whose principles and properties are described.
• Implementation through dedicated software.
• The solution to the problem.
• Interpreting the solution.
• Analysis of the variability of solutions on the basis of input parameters.

In particular, SilverDecisions and R are used in the classroom. Practice sessions are held during the course, in which students complete activities and go through exercises with their laptops, aimed at the described procedure (identifying a model, implementing data, solutions and sensitivity analysis). These exercises are used as self-assessment of learning of the above-mentioned aspects.  

All the parts of the exam aims to verify:

• The ability to identify a model in line with the hypothesys theories and data assigned.
• The ability to implement the model with the appropriate software.
• The ability to interpret the software’s output.
• The ability to assess the sensitivity of the solutions compared to the input parameters.

  The exam will consist of three parts, namely: an individual assignment, a written exam, and a group project.  

  1) Assignments: Students will be required to solve two individual assignments, with problems covering the entire syllabus, to be carried out individually with the use of the software, within a given time window specified by the teachers.  

  2) Written Exam: students will have to solve questions and short exercises, without the use of the software. The written exam will be solved individually during an in-presence session scheduled in one of the official dates established by the School.   

   

• The assignments will count for 4/5 of the final grade, the final exam for 1/5.

• Slides and notes prepared by the course teachers 

• James G., Witten D., Hastie T., Tibshirani R.: An introduction to statistical learning, with application to R. Springer, New York 2013. https://www.statlearning.com 

• Hastie, T., Tibshirani, R., Friedman, J.: The Elements of Statistical Learning: data mining, inference and prediction., Springer-Verlag, New York 2009