Détail d'un enseignement

Data science for marketing analytics

ECTS : 6

Description du contenu de l'enseignement :

“Data-driven marketing is the next normal” McKinsey claimed in 2021. From customer segmentation to the analysis of the effects of a marketing campaign, all the way to sentiment analysis on consumer reviews, data can be leveraged in every step of a marketing strategy. But while the use cases for data in marketing are numerous, and the potential for increased value exponential, the challenges faced by the data-driven marketer are also plenty: collecting quality data, cleaning, analyzing and processing those data sources, identifying a machine learning strategy to solve a specific issue, productionizing a data-driven marketing product, etc. This course will provide future young professionals with the tools and mindset to approach any data science problem, and autonomously design and implement an end-to-end project based on marketing data.

SessionTopic
1Introduction to data science in marketing
2Data preparation and visualization
3Exploratory data analysis
4Fundamentals of Machine Learning (1) - train, test, metrics
5Fundamentals of Machine Learning (2) - model explainability
6Regression algorithms - Predict customer satisfaction score
7Classification algorithms - Customer churn detection
8Clustering algorithms - Customer segmentation
9Project : Customer lifetime value
10Natural language processing - Sentiment analysis on customer reviews
11Productionizing a data science project
12Final Exam

Compétence à acquérir :

At the end of this course, students are able to :

Explain the use & value of data science in a marketing context.
Given a specific data source and a general objective, implement a data processing and analysis strategy.
Use insights extracted from raw data to formulate a marketing strategy.
Identify and defend the selection of machine learning algorithm(s) to address a specific problem.
Design and implement an end-to-end project based on marketing data.

Mode de contrôle des connaissances :

In-class assignments 50%
Final exam 50%

The numerical grade distribution will dictate the final grade. The passing grade for a course is 10/20.

Class participation: Active class participation – this is what makes classes lively and instructive. Come on time and prepared. Class participation is based on quality of comments, not quantity.
Exam policy: In the exam, students will not be allowed to bring any document (except if allowed by the lecturer). Unexcused absences from exams or failure to submit cases will result in zero grades in the calculation of numerical averages. Exams are collected at the end of examination periods.

Bibliographie, lectures recommandées :

“The Data Science Handbook: Advice and Insights from 25 Amazing Data Scientists” by Carl Shan, William Chen, Henry Wang, and Max Song, 2015
“The Ultimate Guide to Marketing AI”, Dataiku White Papers, 2021