CAS in Advanced Machine Learning

In many disciplines, the amount of available data and the computing capacity are growing rapidly. This enables the application of machine learning methods on tasks previously being reserved for humans. Trained machines outperform homo sapiens in more and more cognitive tasks. As with other disruptive technology emergences, the resulting automation potential represents a huge benefit for the human society, but also comes with new challenges and risks. This CAS offers the opportunity to complement your data science competences with a formal deepening and broadening of knowledge and skills on machine learning and artificial intelligence. The format is designed to align with the participants main study and or professional activities. The teaching and learning approaches are team and discussion oriented and designed to develop practical competency.

The program is organised into six modules, running over 18 course days from August to January and targets professionals and researchers in the private and public sector. The content covers a review of machine learning methods, established applications, the research frontier and philosophical and ethical aspects. The difficulty is at a university master level and assumes own basic machine learning experience, programming skills and a higher education degree with some mathematical background. The program focuses on concepts and usage of common machine learning tools, not so much on theoretical elaboration of the mathematics, statistics and informatics.

Objectives

Course competence is developed throughout six modules and a CAS project work. On completion the graduates will (be able to):

• design, tune, train and measure performance of neural networks with advanced deep learning libraries
• understand the inner mechanisms of neural networks during training
• familiar with active research in machine learning
• understand and communicate scientific publications on machine learning and artificial intelligence
• familiar with the philosophy and ethics of extended andartificial intelligence
• familiar with new and advanced machine learning models

If there are free places, modules can be attended individually.

Module 1: Review of machine learning, practical methodology and applications 

Block module. Review of basic principles, concepts, practical methodology and applications of machine learning.

Module 2: Deep networks 

Block module. Study of established deep network applications.

Module 3: Advanced Models I 

This year we expect to do time-series and autoencoders. Module 3 traditionally takes place at the mediterranean coast. Full pension hotel accommodation is included in the CAS fee. 

Module 4: Selected topics on machine learning 

Participants study selected publications on machine learning and artificial intelligence and present them to the others.

Module 5: Philosophy and Ethics of extended cognition and artificial intelligence 

Artificial Intelligence as a scientific field dates back to the 1950s. This module concerns key philosophical and ethical questions and discussions triggered by the existence of intelligence outside the human brain. 

Module 6: Advanced Models II

This year we expect to do Transformers and Reinforcment Learning. Module 6 traditionally takes place in the wonderful historic hotel Regina in the ski resort Mürren (Bernese oberland), only about two train hours from Bern. Full pension hotel accommodations are included in the CAS fee. 

All modules

The duration of all modules corresponds to approximately 20 classroom hours each and module work (expected eff ort is 30 hours), with each complete module qualifying for 2 ECTS points. The expected workload for the CAS Project is 120 hours.Main tools and CAS  language are Python, TensorFlow and Git. Other tools may be used, then with limited support. Computational resources are offered.

Information Events

Learn everything you need to know about the CAS in Advanced Machine Learning. One introduction is mandatory, remote participation is possible. 

Introductionary courses

Prepare yourself for the CAS Modules. We offer the following introductionary courses to refresh your knowledge.

Modules

All Information about Modules 1-6. 

Course materials are accessed via the Ilias learning platform upon receiving the password. 

Events and other important dates

Information Events

Learn everything you need to know about the CAS in Advanced Machine Learning. One introduction is mandatory, remote participation is possible. 

Introductionary courses

Prepare yourself for the CAS Modules. We offer the following introductionary courses to refresh your knowledge.

Modules

All Information about Modules 1-6. 

Course materials are accessed via the Ilias Learning Platform.

Events and other important dates

The Certificate of Advanced Studies (CAS) in Advanced Machine Learning (AML) is offered by the Mathematical Institute.

Program management

• Prof. Dr. Jan Draisma 
• Prof. Dr. Tobias Hodel
• Prof. Dr. Paolo Favaro
• PD Dr. Sigve Haug (director of studies)
• Prof. Dr. Christiane Tretter
• Prof. Dr. Thomas Wihler (chair)

Lecturers            

Lecturers include

• Prof. Dr. Dr. Claus Beisbart 
• PD Dr. Sigve Haug 
• Dr. Kinga Sipos 
• Dr. Mykhailo Vladimirov 
• Dr. Geraldine Schaller Conti
• Dr. Lorenzo Brigato
• et al. 

The CAS is at a university master level and programming and some machine learning skills from eduction or profession are required, e.g. the CAS Applied Data Science.

Target groups

Aimed at students and professionals from the public and private sector that hold a degree from a university or a university of applied sciences (e.g. BSc, MSc, PhD).

SUITABLE FOR MANAGEMENT ► wanting to know how machine learning is performed, limitations and possibilities, ethical aspects

RELEVANT FOR DATA ANALYSTS ► wanting to deepen and update their machine learning skills

APPLICABLE TO CONSULTANTS ► with a desire to know and exploit the possibilities offered by machine learning methods

INTENDED FOR RESEARCHERS ► wanting to extend the machine learning application in their field

Standard data sets are provided, but participants are encouraged to bring or acquire their own. If you have any questions regarding whether this program could work for you, please do not hesitate to contact us.

Admission requirements

• A university degree and basic knowledge of mathematics, statistics, programming, machine learning and professional or research experience in the field of data analysis are required for admission to the course. The required basic knowledge is based on the level of an introductory lecture as part of an undergraduate master’s degree. The program management specifies these requirements.
• Exceptions to the admission requirements can be approved by the program management “sur Dossier”. For people without a university degree, they can impose additional requirements for admission to ensure that they can successfully complete the course.
• Interested parties who only want to take part in individual modules can be admitted, provided that there are free course places.

The program management decides on admission to the course. There is no entitlement to admission.

Registration opens in December and a maximum of 20 registrations can be accepted each year. Registrations are processed in the order of arrival. The CAS can only be offered if there are sufficient registrations by the deadline.

Deadline: 2025-06-30

Program fees

Regular CAS program: CHF 9'600.-

Students currently enrolled in an University or University of Applied Sciences: CHF 5'600.-

Employees of University of Bern: CHF 5'600.-

lnclusive of all modules, performance assessments, certificates, materials & teaching platforms, coffee breaks, full pension hotels (Module 3 and Module 6) and diploma apero.

If there are free places, modules can be attended individually. Prices are CHF 300.- per half day. Individual modules are accredited with certificates which are accumulated for the full CAS AML

Random examples of what CAS participants say:

"I can better support customers in Machine Learning projects and have become more efficient in the implementation."

"The applied work after each module meant lots of coding and some swearing, but the good kind, where you are finally satisfied by what you achieved."

"I would like to thank Sigve Haug for organizing and leading the very insightful course “Certificate of Advanced Studies in Advanced Machine

Learning AML."

”

Two random examples of what CAS participants did:

A CAS AML project on Natural Language Processing, classifying and simplifying documents for dyslexics.

A CAS AML project publication here which used transformers for patent analysis.