Nadine Kupfer
Seminar: ‘Applied AI in Practice’
In order to make targeted use of the potential of AI within a business, knowledge of machine learning, deep learning and generative AI is required. This seminar provides a practical introduction to the technical and methodological foundations needed to understand, develop and perform assessment of AI solutions.
The main programming language used in the seminar is Python, the standard language for data science and AI. Participants will learn how to use AI coding assistants and coding agents to implement and test their own applications. No prior programming experience is required: participants will learn the basics of Python through self-study via an online course.
Data forms the basis of all AI: participants will analyse and visualise datasets, assess their quality and apply them to business use cases.
Participants will learn about machine learning and the key learning paradigms: supervised, unsupervised and reinforcement learning. They will structure their projects using the CRISP-DM and CRISP-ML(Q) process models. Using practical examples, they learn to train their own models for classification, regression and clustering in Python and to perform assessment of their results.
In the area of deep learning, participants learn how neural networks are structured and trained, where their limitations lie, and how their results can be explained (Explainable AI). For image classification, they implement Convolutional Neural Networks (CNNs). Through autoencoders, they become familiar with another Architecture and its areas of application. Examples from industrial practice demonstrate where computer vision is used.
The course concludes with generative AI. Participants will understand how Large Language Models (LLMs) based on the Transformer Architecture work, and will utilise them programmatically via APIs. Using embeddings, they will implement semantic search and Retrieval-Augmented Generation (RAG). They will also learn to apply fine-tuning as a further approach to customisation.
All content is accompanied by hands-on exercises and practical application examples. The aim of the seminar is to gain a technically sound understanding of modern AI methods and to prototype one’s own AI applications.
Key facts at a glance
What you get
After attending the seminar, students will be able to
set up a Python development environment and use Python, with the support of coding assistants, to implement and test machine learning and deep learning applications in practice.
use various types of data for business applications, analyse and visualise datasets exploratively, and assess their quality.
select the appropriate machine learning paradigm for specific problems, structure ML projects in line with established process models, and apply classification, regression and clustering methods in a targeted manner during assessment of their results.
Explain the architecture and training of neural networks, and assess the limitations and interpretability of deep learning models.
Implement convolutional neural networks in Python and apply them to image classification; explain the functioning and areas of application of autoencoders; and assess the potential applications of computer vision in industrial use cases.
Explain how LLMs based on the Transformer architecture work; utilise LLMs programmatically via APIs; use embeddings for semantic search and RAG; and classify fine-tuning as a customisation approach.
The seminar covers content from Module 3 of the ‘Foresight AI’ higher education certificate, which is currently being developed as part of the ‘TCW@ForesightAI’ funded project.
Entry requirements
There are no entry requirements. As capacity is limited to 15 people, places will be allocated on a ‘first come, first served’ basis.
Language of instruction
German
Seminar fee
Free of charge
Dates & Venues
Sat 31 Oct, 8.15 am–5.20 pm, online
Sat 7 Nov, 8.15 am–5.20 pm, online
Sat 28 Nov, 8.15 am–5.20 pm, online
Sat 5 Dec, 8.15 am–5.20 pm, online
Registration
Important note: As the number of participants is limited, places will be allocated on a ‘first come, first served’ basis. By registering, you are simply expressing your interest in taking part. Registration does not yet constitute a binding confirmation of a place. Once we have received your registration, we will check the available capacity and inform you as soon as possible whether a place can be offered to you.
Contact