Highlights

The consortium partners meet for the first time face to face in Dortmund. Two days of intense workshops ensue.

The consolidated development pan for the Didactic Concept and the Use Case Data Preparation work-packages is now agreed upon. There is much homework…

The backbone of our platform is now ready. Wee will feed the beast in the coming months and years by adding and refining the project content!

As a result of the chosen methodology for identifying related Machine Learning teaching concepts from literature, hundreds of papers were identified as potentially related. The process of reading the associated abstracts and filtering out the non-pertinent papers is now finished! As it turns out, less than 10% of the publications are, in fact, useful to our endeavor.

The first data-set was offered by TRUMPF SE + Co. KG. Use-Case preparation and an exemplary solution pending. Many thanks to our industrial partners for this very interesting and highly relevant data-set!

The full text of the papers marked as relevant for the development of the MachineLearnAthon teaching concept is now read and systematized 🙂

July 27, 2023

At TU Dortmund University, the MachineLearnAthon team experienced a change in personnel. Anne Meyer and Alexandru Rinciog will be leaving TU Dortmund, and Lara Kuhlmann will join the project team. Having already worked with Michal Tkáč on previous projects, she is familiar with our collaboration. We thank Anne and Alexandru for their valuable contributions and wish them all the best for their future paths.

We are delighted to share that Anne Meyer has joined the Karlsruhe Institute of Technology (KIT) as a full professor. While she is leaving TU Dortmund University, she will remain connected to the MachineLearnAthon project as part of our advisory board. We warmly congratulate her on this achievement and look forward to continuing our collaboration in her new role.

In October 2023, the MachineLearnAthon consortium met in Košice for its project meeting hosted by the local partner university. The discussions focused on planning upcoming dissemination activities, organizing future multiplier events, and defining topics for the micro lectures. The meeting provided valuable input for shaping the project’s outreach strategy and educational content, strengthening collaboration among partners as MachineLearnAthon continued to grow.

October 10, 2023

In November 2023, the MachineLearnAthon project was represented by Lara Kuhlmann at an event hosted by the German National Agency DAAD in Bonn. The meeting brought together various Erasmus+ projects for exchange and best practice presentations. Lara enjoyed engaging discussions with other project representatives and gained valuable insights to further strengthen our own activities. The event proved to be both inspiring and helpful for future collaboration and project development.

The MachineLearnAthon team had an inspiring exchange with Prof. Frischemeier from the University of Münster (WWU). The discussion focused on similarities between our projects and explored opportunities for future cooperation. This meeting opened up new perspectives for collaboration and knowledge sharing, reinforcing the interdisciplinary spirit at the heart of MachineLearnAthon.

In January 2024, the MachineLearnAthon consortium submitted its joint paper “MachineLearnAthon: An Action-Oriented Machine Learning Didactic Concept” to arXiv. The publication introduces an innovative teaching format designed to make machine learning education more inclusive and practice-oriented across disciplines. At its core are ML challenges based on industrial datasets, guiding students through the entire pipeline—from data preparation to evaluation—while fostering data literacy and hands-on skills. The concept aims to support reliable and responsible application of machine learning methods in diverse academic contexts.

January 29, 2024

At the Pro2Tech workshop, Artur Schweidtmann from TU Delft presented the MachineLearnAthon project to an audience of around 30 participants. The event was aimed at practitioners and professionals interested in learning more about machine learning and its practical applications. The presentation offered valuable insights into our didactic concept and sparked engaging discussions on how action-oriented teaching formats can help bridge the gap between theory and industrial practice.

In March 2024, the MachineLearnAthon consortium met in Dortmund for a productive project meeting. The partners focused on defining the structure of the upcoming micro lectures, developing a unified corporate design for all teaching materials, and planning the implementation of the MachineLearnAthon course at their respective universities. The discussions laid important groundwork for aligning content and visual identity across institutions, ensuring a coherent and engaging learning experience for students.

In April 2024, the MachineLearnAthon project was presented in a lecture on demand forecasting at TU Dortmund University. The lecture, held by Lara Kuhlmann as part of the series Unternehmenslogistik und Supply Chain Management, introduced students to the project’s goals and teaching concept. The presentation connected machine learning applications with practical logistics and supply chain topics, sparking interest in interdisciplinary approaches to data-driven decision-making.

April 24, 2024

An industrial company from the Ruhr area agreed to share its historical sales data with the MachineLearnAthon consortium for use in upcoming challenges. To formalize this collaboration, both parties will sign a data transfer agreement, granting the project the right to publish and utilize the dataset for educational and research purposes. This partnership enriches our collection of real-world use cases and strengthens the link between academia and industry within MachineLearnAthon.

In June 2024, the MachineLearnAthon team submitted its paper “Teaching Machine Learning to Programming Novices: An Action-Oriented Didactic Concept” to the Interdisciplinary Information Management Talks (IDIMT) Conference. With over 30 years of history, IDIMT has established itself as an international forum for interdisciplinary exchange on software-intensive systems, information and knowledge management, social media, and business engineering. The submission marked an important step in sharing our didactic concept with a multinational audience and contributing to ongoing discussions on innovative approaches to teaching machine learning across disciplines.

MachineLearnAthon announced its cooperation with the Graduate School of Logistics to host this year’s PhD Winter School on Predictive Logistics. Taking place in Dortmund from December 2–6, doctoral students and nine international scientists will explore machine learning methods for forecasting and business analytics. The event uniquely combines scientific insight with practical application through live case studies, interdisciplinary teamwork, and networking opportunities. Offered as a multiplier event within MachineLearnAthon, the Winter School promises an inspiring mix of research exchange, hands-on learning, and festive Westphalian traditions.

July 31, 2024

At the end of August 2024, the MachineLearnAthon consortium gathered in Delft for its project meeting hosted by Delft University of Technology. The partners discussed progress on the acquired industrial use cases, finalized topics for upcoming micro lectures and tutorials, and explored possibilities for continuing the collaboration beyond the current project phase. Productive discussions and inspiring ideas made this meeting an important step toward shaping the next stages of MachineLearnAthon.

In September 2024, Michal Tkáč represented the MachineLearnAthon project at the IDIMT Conference in Hradec Králové, Czech Republic. He presented the paper “Teaching Machine Learning to Programming Novices: An Action-Oriented Didactic Concept”, sharing insights into inclusive and practice-oriented approaches for teaching machine learning to students with limited programming experience. The presentation enabled dissemination of our work beyond consortium countries and encouraged international dialogue on innovative ML education methods.

In October 2024, the course MachineLearnAthon – Machine Learning for Beginners started at TU Dortmund University. Taught by Ina Dormuth and Lara Kuhlmann, the elective master’s course welcomed ten participants, including three women. The students come from the fields of logistics and industrial engineering, bringing diverse perspectives to the learning environment. The course introduces fundamental machine learning concepts through hands-on challenges, fostering interdisciplinary collaboration and practical understanding.

October 17, 2024

The company Trumpf agreed to our data transfer agreement and kindly provided production data from its metal sheet extraction process. This dataset will serve as the basis for new MachineLearnAthon challenges, allowing students to apply machine learning methods to real manufacturing data and explore topics such as process optimization and predictive analysis. The collaboration highlights the valuable support of industry partners in enriching our educational materials with authentic use cases.

This week, the PhD Winter School on Predictive Logistics took place in Dortmund — this year organized in cooperation with MachineLearnAthon. Over five inspiring days, doctoral students and international researchers explored machine learning methods for forecasting and business analytics. Participants attended insightful talks, discussed topics such as trustworthiness and bias in ChatGPT, and collaborated in interdisciplinary teams on a live case study focused on sales data forecasting. The event combined academic depth with practical relevance and created an engaging atmosphere for learning, exchange, and networking.

The MachineLearnAthon team announced its participation at the LearnTec fair in Karlsruhe from May 6–8, 2025. The consortium looked forward to engaging with visitors, exchanging ideas, and presenting the latest project results to an international audience.

January 14, 2025

We were truly inspired by the results of the students from the MachineLearnAthon – Machine Learning for Beginners course at TU Dortmund University. The participants developed intriguing methods and achieved impressive outcomes in their machine learning challenges. Their constructive feedback was equally valuable — including suggestions to define expectations more precisely, which we will incorporate into the next course run in the summer term. The experience highlighted how collaborative learning continues to strengthen and refine our didactic concept.

In March, the first MachineLearnAthon challenge was published: Hand-drawn Unit Operation Recognition. The dataset, provided by the Process Intelligence Research group at Delft University of Technology, focuses on classifying hand-drawn unit operation symbols in chemical engineering. This advanced task invites participants to develop machine learning models that contribute to the digitization of Process Flow Diagrams and improve accessibility for engineers.

Two new datasets were published within the MachineLearnAthon project: Blueberry Production and Fiber Content in Wheat and Barley Grain Crops. Provided by the team from the University of Tartu Institute of Computer Science, these intermediate-level classification tasks focus on predicting agricultural outcomes based on variables such as irrigation, soil quality, weather data, and crop composition. Both datasets support developing machine learning models that help optimize production processes and improve yield predictions in modern agriculture.

April 29, 2025

The MachineLearnAthon team participated in the LearnTec fair for digital education in Karlsruhe. Together with the EXPERTISE Erasmus+ project, the consortium presented innovative learning materials on Machine Learning and Digital Skills. Engaging discussions and lively exchanges made this event a great opportunity to share project results with a broad audience.

Last month, the MachineLearnAthon consortium met in the beautiful city of Tartu for its latest project meeting. The partners discussed progress, outlined upcoming steps, and exchanged ideas on developing innovative and engaging Machine Learning teaching materials. Alongside productive sessions, there was also time to reconnect and enjoy some social activities together.

The TRUMPF dataset, previously validated twice with student groups from TU Dortmund University, has now been made publicly available on Kaggle. This dataset provides valuable real-world industrial data for machine learning applications and serves as the foundation for one of the MachineLearnAthon challenges. By sharing it openly, we aim to encourage broader engagement with practical ML problems and support collaborative learning across institutions.

July 31, 2025

Another successful term of the MachineLearnAthon – Machine Learning for Beginners course at TU Dortmund University came to an inspiring conclusion. During their final presentations, students showcased impressive solutions to the TRUMPF challenge, demonstrating creativity, dedication, and strong analytical skills. We were delighted to share their results once again with our contact person at TRUMPF — continuing a valued collaboration that bridges academic learning with real-world industrial applications.

In September 2024, the Master’s course Explainable and Automated Machine Learning started at the University of Tartu with around 40 participants. Educational materials developed within the MachineLearnAthon project were integrated into the course, supporting students in exploring key concepts of transparency and automation in machine learning. This successful implementation highlights how our project outcomes are being adopted across universities to enhance interdisciplinary ML education.

In October, the MachineLearnAthon consortium hosted a multiplier event. Throughout the week, PhD students, postdocs, and Prof. Artur Schweidtmann from the Process Intelligence Research group participated in a hackathon focused on applying AI in chemical engineering. The event provided valuable hands-on insights into integrating domain knowledge with modern machine learning workflows.

October 24, 2025

In November, the MachineLearnAthon consortium met for its final project meeting in the beautiful High Tatras Mountains, Slovakia. Warmly hosted by Michal Tkac and Miroslava Barkóciová, the partners finalized project outcomes and prepared the concluding report as MachineLearnAthon approaches its completion at the end of December. The meeting was both productive and inspiring. Alongside fruitful discussions, participants enjoyed a winter wonderland experience with mountain views and a visit to the impressive Tatranský ľadový dóm Ice Palace.

The MachineLearnAthon micro lectures and tutorials have been uploaded to our official YouTube channel and are now publicly available to everyone. These short, focused videos cover key concepts and practical methods in machine learning, designed to make the topic accessible for learners from different academic backgrounds. By sharing our materials openly, we aim to support self-paced learning and encourage wider engagement with the MachineLearnAthon didactic concept.