Data science and machine learning

MGT-502 / 5 credits

Teacher(s): Thurm Boris, Vacat ., Vlachos Michail

Language: English

Withdrawal: It is not allowed to withdraw from this subject after the registration deadline.

Summary

Hands-on introduction to data science and machine learning. We explore recommender systems, generative AI, chatbots, graphs, as well as regression, classification, clustering, dimensionality reduction, text analytics, neural networks. The course consists of lectures and coding sessions using Python.

Content

Basic regression methods. Predicting numeric values.
Basic classification methods. Predicting categorical values: logistic regression, k-NN classification, decision trees, random forests
Fundamental concepts: cost function and optimization, gradient descent, K-fold cross-validation, overfitting, model calibration, confusion matrix
Dimensionality reduction: Principal Component Analysis, Multidimensional Scaling, non-linear dimensionality reduction (ISOMAP, t-SNE), curse of dimensionality
Clustering
Neural networks and Deep Learning
Text analytics: text representation, sentiment classification, similarity search
Recommender Systems
Graph Analytics: Pagerank, centrality, 6-degrees of separation
Generative AI and chatBots. chatGPT.

Keywords

AI, Machine learning, Data Science Algorithms, Recommender Systems, Graphs, Generative AI, Regression, Classification, Dimensionality reduction, Clustering, Neural networks, Text analytics, Python

Learning Prerequisites

Required courses

Statistics and data science (MGT-499)

Important concepts to start the course

Basic Probability and Statistics knowledge (random variables, expectation, mean, conditional and joint distribution, independence, Bayes rule, central limit theorem)
Basic linear algebra (matrix/vector multiplication, system of linear equations)
Multivariate calculus (derivative w.r.t. vector and matrix variables)
Basic programming skills (Python)

Learning Outcomes

By the end of the course, the student must be able to:

Describe the principal types of machine learning algorithms
Investigate data, data types, and problems with the data
Choose an appropriate Machine Learning method for a given task
Implement Machine Learning algorithms in Python
Optimize the main tradeoffs such as overfitting and computational cost vs accuracy
Conduct a Data Science project

Transversal skills

Plan and carry out activities in a way which makes optimal use of available time and other resources.
Demonstrate the capacity for critical thinking
Access and evaluate appropriate sources of information.
Use a work methodology appropriate to the task.

Teaching methods

Lectures
Lab sessions: coding exercices
Data Science projects

Expected student activities

The students are expected to:

attend lectures and lab sessions;
work on the weekly theory and coding exercises;
complete assignments (graded);
conduct data science projects making use of the theory learned during lectures and code developed during lab sessions (graded)

Assessment methods

Quiz: 40%
Coding assignments: 30%
Group Project: 30%

Supervision

Office hours	No
Assistants	Yes
Forum	No
Others	Slack channel

Resources

Virtual desktop infrastructure (VDI)

Bibliography

[not mandatory] Data Science for Business: What You Need to Know about Data Mining and Data-Analytic Thinking, by Foster Provost and Tom Fawcett

Ressources en bibliothèque

Data Science for Business / Provost & Fawcett

Notes/Handbook

Slides will be made available on the course Moodle page. Notebooks will be made available in a GitHub repository.

Moodle Link

https://go.epfl.ch/MGT-502

In the programs

Semester: Spring
Number of places: 40
Exam form: During the semester (summer session)
Subject examined: Data science and machine learning
Lecture: 3 Hour(s) per week x 14 weeks
Exercises: 2 Hour(s) per week x 14 weeks

Reference week

	Mo	Tu	We	Th	Fr
8-9
9-10
10-11
11-12
12-13
13-14
14-15
15-16
16-17
17-18
18-19
19-20
20-21
21-22

Légendes:

Lecture

Exercise, TP

Project, other