|Zeiten:||Zweiwöchentlich, 18:00-20:30, Tag tbd|
|Zeitraum:||3.10.21-30.09.22, Ferien und September ausgeschlossen|
|Kosten:||45€ pro Monat (siehe Kosten)|
[Je nach Wunsch der Teilnehmer, findet der Unterricht auf Deutsch oder auf Englisch statt.]
In the last years and decades results from the areas of artificial intelligence and machine learning became ubiquitous in the world surrounding us. Some of us use them every day, for example when we translate a text, use snapchat filters or speak with artificial assistants. Also the more fun and less applied capabilities of large models based on neural networks are phenomenal. Reinforcement learning based programs beat the world champions in go (alphaGO) and dota, generative models can create images from text (dalle), synthesize new faces (stylegan), even write entire songs with lyrics, instruments and voice.
At the same time, many of the ideas involved in such breakthroughs are not too difficult to understand even with little prior knowledge. Moreover, thanks to the work of thousands of software engineers, we now have access to amazing open source libraries and data sets that allow virtually anybody with enough time and dedication to design, build and test their own models and ideas.
In this course we will leverage the modern AI software ecosystem to give you an introduction into the foundations of modern AI, a glimpse into what is possible today and ideas for what may be possible soon.
What will we learn?
We will begin the course with a short introduction to the methods from programming, analysis, statistics and linear algebra needed to understand the basics of artificial neural networks. Then we go on to explore different architectures and designs, from basic classification and regression to convolutional neural networks, transformers, GANs and so on. There is a wealth of topics that could be explored, exactly which we end up choosing is in part up to the students.
How will we work?
A profound understanding of the topics treated in this course will require broad knowledge in mathematics, theoretical AI and programming. Unfortunately, it would be impossible to treat on all of these fields in depth in the course’s format. Instead, the lectures will be a mixture of theoretical background and hands on exercises/experimentation in a preconfigured programming environment. In the hands-on parts, the students will modify, inspect and execute code, thereby gaining an intuitive understanding of the algorithms they are dealing with. Depending on individual preferences, the students will pick from exercises going more into programming, theory or application and eventually share the obtained results with each other.
What are the goals?
At the end of the course the students will have a broad overview of the foundations, implementation and applications of modern artificial intelligence. This will make it easier for them to dive deeper into selected topics on their own and will also enable them to read and understand many of the developments that are to come.
Haben Sie Fragen?
Rufen Sie uns gerne gleich an oder schreiben Sie eine E-Mail:
0151 701 66 162 | firstname.lastname@example.org