CAS Foundations of AI and Machine Learning

The CAS Foundations of AI and Machine Learning is designed for professionals with a technical or scientific background and several years’ professional experience who wish to acquire or deepen their expertise in artificial intelligence and machine learning. The target audience includes:

• Data scientists
• Data or business analysts
• Software engineers and developers
• Specialists in analytical marketing
• Engineers, scientists and technical professionals with an interest in AI
• Individuals responsible for AI and innovation projects
• IT project managers and consultants

The CAS Foundations of AI and Machine Learning focuses on the following questions:

• How can one create optimal conditions for machine learning?
• How do modern deep learning models and generative AI work, and where are they successfully applied?
• How are large language models (LLMs) and generative AI changing the processing and use of language? What are their opportunities and limitations?
• How do systems learn through interaction with their environment, and in which use cases is reinforcement learning applied?

Students will acquire both theoretical foundations and practical skills in the following areas:

• Machine learning
• Deep learning and neural networks
• Language AI and large language models (LLMs)
• Reinforcement learning and autonomous decision-making

Module "Machine Learning"

Learning Objectives

• Participants will be familiar with the fundamental concepts and areas of application of AI and will have a general understanding of key developments.
• They will be familiar with the essential fundamentals and best practices for the use of ML methods
• They will be able to select a suitable ML method for a given dataset and prepare the features accordingly
• They will be able to develop self-learning scripts using ML algorithm libraries such as Python/sklearn

Contents

• Fundamentals, historical development and classification of Artificial Intelligence (AI)
• Fundamentals, application concepts and best practices for Machine Learning (ML)
• Selected Machine Learning algorithms (clustering, classification, anomaly detection)
• Feature engineering 

Module "Deep Learning"

Learning objectives

• You will be familiar with and understand the fundamentals and relevant architectures of deep learning
• You will be familiar with the latest developments in deep learning
• You will be able to independently apply suitable deep learning methods to new problems using the framework presented in the lecture and used in the practical session

Contents

• Fundamentals of deep learning (Structure and function of neural networks)
• Training, optimisation and regularisation of deep learning models
• Modern architectures such as convolutional neural networks, recurrent neural networks and transformers, and current developments
• Applications such as object recognition and segmentation, and the evaluation of deep learning models

Module "Language AI Introduction"

Learning objectives

• You will be familiar with the key methods of language AI, including large language models (LLMs)
• You will be able to assess how well an AI-based solution might perform for a specific task involving text as the modality
• You will be able to implement simple language AI systems and evaluate their quality

Contents

• Introduction to modern language AI and large language models (LLMs)
• Evaluation of systems that primarily use AI-based text processing
• Practical applications such as text classification, sentiment analysis, topic modelling and chatbots

Module "Reinforcement Learning"

Learning objectives:

• Participants will understand how RL can be used to optimise control and regulation processes.
• They will be able to identify RL methods for optimising production processes and for autonomous decision-making on the basis of specific problems.
• They will be able to develop adaptive control strategies using real-world datasets or simulations.

Contents

• Introduction to Reinforcement Learning, including Deep RL
• Control and regulation through sequential decision-making processes: value functions and exploration-exploitation
• Sampling-based methods: temporal difference learning, Q-learning
• Policy gradient methods

The programme includes a range of activities, such as lectures, practical exercises and case studies, group work and self-study (preparation and follow-up).

Module tests

The CAS Foundations of AI and Machine Learning has four modules. Classes take place once a week on Tuesdays from 9.00 to 17.00 (8 lessons). Each day’s sessions are divided into two blocks of 4 lessons each, with each block comprising 2 lessons of theory and 2 lessons of practical work. During the practical sessions, participants consolidate what they have learnt using specific examples, which they work on using the relevant software on their own laptops.

You can find the timetable here.

The lecturer team consists of recognised experts with expertise in both academic and practical fields. Here is an extract from the list of lecturers:

• Prof. Frank-Peter Schilling, CAI
• Dr. Don Tuggener, CAI
• Dr. Philipp Denzel, CAI
• Dr. Jorge Pena Queralta, CAI
• Dr. Lilach Goren Huber, IDS
• Dr. Nima Riahi, IDS

Admission requirements for applicants with a higher education qualification

Admission to the course requires:

• A qualification (diploma, licentiate, bachelor’s or master’s degree) from a state-recognised higher education institution or one of its predecessor institutions.
• Two years’ relevant professional experience at the start of the further training programme.

Admission requirements for applicants without a higher education qualification

Admission to the course requires:

• Proof of a qualification in higher vocational education (Tertiary-B): Vocational Examination (BP) (Federal Certificate), Higher Vocational Examination (HFP) (Federal Diploma) or Higher Vocational School (HF). In exceptional cases, other candidates may be admitted if their ability to participate is demonstrated by other evidence.
• 3 years’ relevant professional experience at the start of the further training programme.
• Successful completion of an admissions interview

February classes: Lessons are generally held in German, with one module in English.
September classes: Lessons for all modules are held in English.

We do not keep waiting lists or offer seat reservations.
Should a place become available in the preceding class, we will allocate it in the order in which applications were received.