Introduction to Deep Learning (Sommersemester 2020)

Lecturer: Prof. Dr. Christoph Lippert (Digital Health - Machine Learning) , M.Sc. Aiham Taleb (Digital Health - Machine Learning) , Jana Fehr (Digital Health - Machine Learning)

General Information

• Weekly Hours: 4
• Credits: 6
• Graded: yes
• Enrolment Deadline: 06.04.2020 - 22.04.2020
• Teaching Form: Lecture / Seminar
• Enrolment Type: Compulsory Elective Module
• Course Language: English
• Maximum number of participants: 98

Programs, Module Groups & Modules

Description

Lecture starts 04/27/2020
Technical onboarding Thursday 04/23/2020 via Zoom

This course is designed to give students an in-depth introuction to deep learning and review the state of the art in important applications, such as computer vision and language modeling. The lectures and exercises are designed around use cases and will use real-world data to gain practical experience with machine learning models and algorithms. The course will start with an introduction to the basic concepts of machine learning and empirical data analysis, such as model fitting, selection and validation. During the second part of the course, we will discuss supervised machine learning, starting with linear models, to non-linear models, including deep neural networks, convolutional neural networks and sequence models.  During the third part of the course, we will discuss unsupervised learning, generative models and self-supervision.

Learning Objectives:

• Understand concepts, methods and algorithms in machine learning
• Ability to empirically analyze real-world data
• Ability to assess the quality and validity of a machine learning model for a given analysis
• Ability to select, develop, implement and apply appropriate machine learning models and algorithms for a given use case.

Requirements

In order to grasp the mathematical background of deep learning algorithms, we expect that students have knowledge in calculus and linear algebra. Thus, for HPI students it is mandatory to have passed Mathematik 2.

Additional knowledge in probability (e.g. from Mathematik 3 - Stochastik) is an advantage, but not mandatory.

Literature

This course will be based on the material provided at http://d2l.ai/

Learning

Online tools include Zoom for Meetings, A Learning Platform (OpenHPI or Teletask or Moodle),
and Google Colab (https://colab.research.google.com/) as programming environment.
For this it will be necessary to use Zoom and Google Services and to install appropriate software.

Examination

Processing of regular exercise sheets, programming project assignments und written final exam

Dates

• Lecture #1:  Monday 15:15 -16:45pm in HS 2
• Lecture #2: Thursday 13:30-15:00pm in HS3
• Tutorials: Thursday 15:15-16:45pm in H2.57

Lecture starts 04/27/2020
Technical onboarding Thursday 04/23/2020 via Zoom

The course is now online on Moodle (Link: https://moodle2.uni-potsdam.de/course/view.php?id=24365, Key: DeepLe@rning),
you can find the links for joining the lecture on Zoom there under 'Administrative announcements.

Zurück