COURSE DESCRIPTION

1. Course specifications

Professor: Dr. Kristof Coussement – Professor in Business Analytics

IESEG School of Management, Lille, France

www.kristofcoussement.com 

Course outline

The course is organized in 5 face-to-face sessions with mandatory lecture + lab time, and optional individual guidance and feedback possibility.

Lecture + Lab
5 days – 4 hours/a day

Individual guidance/feedback
5 days – 2 hours office hours

2. Aim of the course

The aim of the course is to introduce in a hands-on way its participants to the data science field. The course is decomposed in three large blocks. First, the course starts by introducing the importance of data science and analytics into our today’s society, discussing various real-life examples and giving an overview of recent academic developments. Second, we introduce a framework that helps the participants to understand the various method typologies and to choose the correct method typology for solving their research problem. A last block of the course is to introduce the participants to the fundamentals of building predictive data science pipelines, including (business) problem definition, timeline construction, data preparation and featurization, data preprocessing methods, experimental setup design (single split versus cross-validation), statistical and machine learning algorithms and their hyperparameter tuning process, evaluation metrics, and visualization methods.

Note 1: the analytical software used throughout the entire course is the graphical user interface of Dataiku or similar, so the participants do not need any programming experience.

Note 2: the course is delivered using interactive lecture time blended with practical lab time.

Note 3: the professor offers a voluntary programming trajectory on the self-learning e-platform Datacamp.com. For participants willing to learn how to program in Python or R, learning Python or R falls outside the scope of the course.

3. Prerequisites of the course

Fundamentals in Statistics
Laptop with DataIku pre-installed: www.dataiku.com/product/get-started

4. Learning Objectives

At the end of the course, the participant is able to:

• spot analytical opportunities for her/his academic field.
• identify the correct method to solve a particular problem.
• understand the various steps in the predictive data science pipeline.
• apply a data science pipeline on a dataset originating from her/his academic field.

5. Evaluation

The overall course grade will be based on participants’ class participation (20%), a review literature study (30%) and a case study assignment (50%).
The review literature study has the intention to collect, read and summarize academic papers in the participants’ academic field of interest.
The case study assignment has the intention to apply the entire data science pipeline on an empirical dataset originating the participants’ academic field of interest.
Note: further details on all assessments will be given during the first lecture.