The MachineLearnAthon is an Erasmus+ cooperation partnership project introducing a new, challenge-based teaching format centered around machine learning (ML) competitions. Our goal is to make machine learning accessible, engaging, and action-oriented by combining theoretical understanding with hands-on application.
Main Results
At the heart of MachineLearnAthon lie four key outcomes that define our approach to modern ML education:
- Didactic Concept:
Developed on principles of action orientation, constructivism, and problem orientation. This concept empowers learners to engage actively with real-world problems rather than passively consume content. - 8 Real-World Challenges:
Authentic prediction tasks based on publicly available datasets form the backbone of our teaching format. Each challenge includes leaderboard access, motivating participants through transparent progress tracking and friendly competition. - 29 Tutorials:
Step-by-step video units (each about 10 minutes long) guiding learners through practical ML implementation in Python—ideal for self-paced study and skill development. - 55 Micro Lectures:
Concise video lectures (each roughly 10 minutes long) covering ML theory, paradigms, frameworks, and data preparation—providing a solid conceptual foundation for subsequent application.
Together, these elements create an integrated learning experience that connects theory with practice and enables independent learning across different levels of expertise.
Our Goals
The project aims to:
- Advance awareness and skills in machine learning across disciplines.
- Enhance data literacy among students with varying levels of programming experience.
- Promote ethically sound, trustworthy, and robust ML solutions.
- Foster international cooperation between students and educators.
- Strengthen collaboration between academia and industry through real-world challenges.
Further Resources
To support educators and institutions in implementing this concept, we provide several resources:
- A white paper, published on arXiv titled MachineLearnAthon: An Action-Oriented Machine Learning Didactic Concept, detailing our didactic framework.
- An extended conference paper, Teaching Machine Learning to Programming Novices: An Action-Oriented Didactic Concept, offering further insights into our methodology.
- A comprehensive course handbook, including examples of how to integrate MachineLearnAthons into university curricula.
What You’ll Find on This Website
Explore our:
- Educational Material: Resources designed for individual learning, including:
– Micro Lectures: Short, focused lectures accompanied by a glossary that helps you find specific topics across all micro lectures.
– Tutorials: Step-by-step guides for hands-on learning, complemented by a comprehensive Tool List comparing different tools for ML implementation. - Use Cases: Eight real-world datasets serving as foundations for our challenges
- News: Monthly insights into ongoing developments within MachineLearnAthon
- Consortium: Learn more about our international partners driving this initiative
These results together illustrate how MachineLearnAthon contributes to the democratization of machine learning—empowering learners from all disciplines to understand, apply, and critically reflect on ML methods in practice.