NTNU Norges teknisk-naturvitenskapelige universitet
  Fakultet for informasjonsteknologi, matematikk og elektroteknikk > Institutt for datateknikk og informasjonsvitenskap

TDT-76 - Deep Learning, høst 2018


TDT-76: Deep Learning (a theory module)

This is the instructor's page for TDT-76.

Course Materials

One book: "Deep Learning (2016)", Goodfellow, Bengio and Courville, MIT Press.

Book Chapters:
  • Chapters 1-5. (General Background)
  • Chapters 6, 7, 8, 9, 10, 11, 12, 14, 15. (Deep Learning in Detail)

The point is not to understand every detail of every chapter, but rather, to absorb the essence of each. This does not mean that you should just skip all the mathematics, but try not to get too bogged down in it. If you don't understand a complex mathematical or algorithmic section, skim it and try to come back to it later if it turns out to be essential for understanding many other parts of the book.

Composer Project Download

Lecture Slides

Old Exams (with Answers)

Other Materials

Primary Instructors

  • Helge Langseth (helge.langseth@ntnu.no)
  • Keith Downing (keithd@ntnu.no)

Course Assistant

Eliezer de Souza da Silva (eliezer.souza.silva@ntnu.no)

Course Meetings

Where: Room 424 ("Vembi") in building "Høgskoleringen 3"

When: Fridays kl 14:15 - 16:00 (throughout the semester)

FIRST MEETING: September 7th, 2018

Final Exam

  • Exam Dates: Wednesday - Friday, November 28-30, 2018
  • Due to the overwhelming popularity of this module, we are forced to spread the evaluation out over 3 days. Each student will receive information from IDI as to the exact day of their exam.
  • On your exam date, you will meet up to both a) demonstrate your project, and b) take a short oral exam (on the theoretical aspects of the course). Results of both activities will be combined to form the final grade.

  • The oral exam will not focus heavily on the most complex linear algebra and calculus associated with neural networks, but a basic understanding of the primary mathematical and computational concepts (e.g., tensors, partial derivatives, objective functions, etc.) and their use in Machine Learning (in general) and Deep Learning (in particular) will be necessary to properly prepare for the exam.
Redaktør: Kontorsjef: Eivind Voldhagen  Kontaktadresse: Audun Liberg  Sist oppdatert: 25.10.2018