| Seminare.EIM: Introduction to Deep Learning (DSBS, CSMS, IIWMS, TMBS, IMPICS) |
| Semester: |
| SoSe 24 |
| Course type: |
| Seminar |
| Lecturer: |
| Dr. rer. nat. Pradeep Banerjee |
| Description: |
| Deep Learning is one of the most vibrant areas of modern machine learning, offering one of the most promising routes to advancing Artificial Intelligence (AI). Deep Learning systems are reshaping the AI landscape across various fields, including language comprehension, speech and image recognition, and autonomous driving. This seminar covers deep neural networks basics and their applications in various AI tasks. We will explore several key paradigms related to expressivity, optimization and generalization properties of modern deep learning systems. Students will gain proficiency in Deep Learning, enabling them to apply it to different scenarios and comprehend current literature in the field. |
| Participants: |
| This seminar is aimed at all Bachelor- and Master- level students in the Informatik and the Techno-Mathematik courses. A maximum of 12 students can participate in the seminar. |
| Pre-requisites: |
| As a prerequisite, this seminar will assume familiarity with basic calculus, linear algebra, and probability. Familiarity with a programming language such as Python is desirable. |
| Learning organisation: |
| The seminar is divided into six blocks (following an introductory session), each lasting two weeks. Every block consists of the following components:
* Week 1: Preparation of a presentation using prescribed sources (book chapters, video lectures, scientific articles).
* Week 2: Presentations by 2 participants, each lasting 25 minutes based on a topic assigned to each participant in the first session of the seminar. |
| Area classification: |
| Studiendekanat Elektrotechnik, Informatik und Mathematik |
| Stud.IP informationen about this course: |
Home institute: Studiendekanat Elektrotechnik, Informatik und Mathematik (E)
Participating institute: Institut für Data Science Foundations (E-21)
Registered participants in Stud.IP: 11
Documents: 5
|
