• Title in French: Projet data
• Course code: tba
• ECTS credits: 3
• Teaching hours: 60h
• Type: advanced course
• Language of instruction: French
• Coordinator: tba
• Instructor(s): Alexandre Chirié (Mantiks), Maximilien Defourné (Mantiks)
• Last update 27/08/2021 by C. Pouet

The course consists of a theoretical part and a practical part, simulating a business project.

• Understand the workflow of a data science project in a business context
• Be able to account for business (collection of needs, project lifecycle, communication) and technical (data, machine learning, scaling) constraints

1. Data science in business
   • The main issues
   • Examples of data project
2. Starting a data science project
   • The constraints of data science projects
   • Finding data
   • Acquiring information
   • Playing with data
3. Lifecycle of a project
   • The Bias-Variance tradeoff
   • Feature Selection
   • Feature Engineering
   • Defining a metric
4. The basic models
   • Regressions (linear, polynomial, penalized et logistic)
   • Decision trees (random forest and gradient boosting)
5. Focus Natural Language Processing (NLP)
   • Word Embedding
   • Example: Sentiment analysis

Check the availability of the books below at Centrale Marseille library.

• Zeng, A and Casari, A. Feature Engineering for Machine Learning. O'Reilly Media.
• Müller, A. and Guido, S. Introduction to Machine Learning with Python. O'Reilly Media.