Welcome to the Introduction to Deep Learning course offered in WS22-23.
General Course Structure
The lecture will be held in person the exercise session will be held online.
In addition, the class will feature weekly exercises consisting of additional videos, practical coding tasks as well as semi-weekly submissions which will be used to determine a bonus for the exam. Please watch the first tutorial video at the beginning of the semester where we will cover the class structure and planning in more detail.
The exercise submissions will start on the first week of the semester. See below, if you are a LMU or TUM student that currently can't sign up for the class regularly via TUM online right now.
External Students & Late TUM Online Registrations
We want to provide access to our lecture for as many students as possible. If you are affiliated with TUM (e.g. LMU student, Ph.D. student, TUM student who cannot register for courses yet but have a TUM token, etc.), we will add you to our class manually. To be added as an external student, please fill in the following form: External student registration form. Note that TUM students can enroll themselves to piazza using a @mytum.de address.
If you are a student from another university or simply
interested in Deep Learning, you can access the lecture and exercises videos through this webpage which will be updated accordingly. All notebooks feature separate tests where you can record your performance, though you will not be able to access to submission website. We will also provide solutions and discussions about the submissions in our weekly exercise videos which are all publicly hosted on this website.
Lecture
See lecture links below.
Lecturer: Prof. Dr. Angela Dai
ECTS: 6
SWS: 4
Tutorial
Submission date: Wednesdays, 15:59
Lead Teaching Assistants: Manuel Dahnert and Yujin Chen and Junwen Huang
Student Teaching Assistants: Dan Halperin, Erik Traise, Haoxuan Li, Wei Cao
For details about the exercises please refer to the first tutorial session. In particular, you will submit your solutions via our submission website (https://i2dl.vc.in.tum.de).
Lecture Slides
- Week 1 - Lecture 1: Introduction to the lecture, Deep Learning, Machine Learning and submission system
- Week 2 - No Lecture, please watch the tutorial/exercise 02 this week below
- Week 3 - Lecture 2: Machine Learning Basics, Linear Regression, Maximum Likelihood
- Week 4 - Lecture 3: Introduction to Neural Networks, Computational Graphs
- Week 5 - Lecture 4: Optimization and Backpropagation
- Week 6 - Lecture 5: Scaling Optimization to large Data, Stochastic Gradient Descent
- Week 7 - Lecture 6: Training Neural Networks
- Week 8 - Lecture 7: Training Neural Networks - Part2
- Week 9 - Lecture 8: Training Neural Networks - Part3
- Week 10 - Lecture 9: Introduction to CNNs (part 1)
- Week 11 - Lecture 10: Introduction to CNNs (part 2)
- Week 12 - Lecture 11: RNN and Transformers
- Week 13 - Lecture 12: Advanced DL topics
Office hours
In addition to the tutorial we also offer to discuss questions in person. This semester, the office hours will be held via TUM-Zoom. For each office hour, we provide a separate zoom-link, see announcement on Piazza. Please keep the following in mind to ensure the office hours remain helpful for everyone:
- Check out the correct zoom link of the time-slot
- Mute your microphone when entering a zoom-session
- Wait patiently until it’s your turn, the TAs will try to answer the questions in the correct order (just like in the usual office hours)
Office hours start on the second week of the class. For one hour, you can have direct contat with one of our TAs to resolve your issues in person. The links to these office hours can be accessed via the forum below and is only available to registered students.
Final Exam and Credits
The final exam date is TBD. The exam will most likely be a traditional onsite exam, so please take this into consideration if you intend to receive credits for this class. Credits are only awarded for students that participated and successfully passed the exam.
As this course is taught every semester, there will be no retake exam; you will have to take next semester's exam (bonus will be transferred).
We provide a mock exam and the solution from the last semester for your reference.
Prerequisites
- Strong mathematical background: linear algebra, calculus.
- Previous knowledge of Python is needed. This includes scripting skills such as operating system calls.
- Machine Learning background is recommended.
Forum
We will use Piazza for discussions, publication of office hour links and to distribute exam related information. We will not be using Moodle. The slides and all material will also be posted and maintained on Piazza, but it will be identical the material hosted on this website.
TUM students can register directly with TUM email address. The other students (e.g. LMU student) need to manually send their email address to us to get enrolled on Piazza.
Contact us
If you have any questions regarding the organization of the course, do not hesitate to contact us at: i2dl@vc.in.tum.de. Please refrain from using the personal email addresses.
For questions on the syllabus, exercises or any other questions on the content of the lecture, we will use the forum discussion board.
