Embedded Files
IFT6759 - Advanced Machine Learning Projects
This is a course offered by Mila and the Université de Montréal in the context of the professional master in machine learning. It can be attended by the master students in computer science as well.
The goal of this course is to prepare master students to work on real machine learning projects. This course is thus designed to connect the theoretical aspects of deep learning with strong practical and coding insights. Having strong theoretical and coding capabilities, in fact, is essential for the students to succeed in their future careers as machine learning developers.
The course is structured in two parts:
Tutorials and frontal lectures (from week 1 to 5)
Project development (from week 5 to week 15)
The first part aims to give the students the basic tools needed to work on real machine learning projects. We will start with some lectures on Linux, HPC cluster usage, git, Colab, PyTorch, deep learning trips and tricks.
In the second part, students will work on real machine learning projects. Students will be assigned to one of the proposed project and they will work in small groups of 2-6 students. During this phase, the progress of the project is constantly monitored by the teacher and the teaching assistants.
This year the projects will be connected to an ongoing open-source machine learning project called SpeechBrain. Before starting the project development phase, we will provide to the students some lectures on speech processing, including tutorials to familiarize themselves with SpeechBrain.
At the end of the course, we expect that students will provide a short report summarizing the main achievements. An oral presentation in front of the teachers and the classmates is required as well.
Evaluation
Students will be evaluated according to different parameters:
Quality of the code.
Quality of the experimental validation.
Quality of the report and oral presentation.
Involvement in the project.
Page updated
Google Sites
Report abuse