{"id":630,"date":"2015-03-19T11:51:18","date_gmt":"2015-03-19T10:51:18","guid":{"rendered":"http:\/\/lwibs01.gm.fh-koeln.de\/blogs\/ciop\/?page_id=630"},"modified":"2015-03-19T12:18:17","modified_gmt":"2015-03-19T11:18:17","slug":"news-4","status":"publish","type":"page","link":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/news-4\/","title":{"rendered":"CIOP News"},"content":{"rendered":"

\n\t

MOPTA (Modeling and Optimization) conference with invited speaker from TH K\u00f6ln<\/a><\/h2><\/div>
By Wolfgang Konen on Juli 11, 2025<\/div>
<\/p>\n
<\/div>\n

<\/a>

\"\"<\/a>

Wolfgang Konen talking about SACOBRA at MOPTA2025<\/p><\/div><\/p>\n

The MOPTA conference (Modeling and Optimization: Theory and Applications, https:\/\/coral.ise.lehigh.edu\/mopta2025\/<\/a>) has been around for many years, since 2001 to be precise. It usually takes place in America, but this year, 2025, European and American scientists met in the middle, so to speak, because it took place in the Azores in June this year.<\/p>\n

In June 2025, optimization scientists from all over the world met on Sao Miguel, the main island of the Azores, to exchange scientific ideas. The research conducted at the Gummersbach campus attracted attention in this international circle, as Wolfgang Konen was invited by Don Jones, one of the co-organizers of the conference, to speak as an invited speaker about his research area of Constrained Optimization<\/em>. This is a special honor, because normally one has to undergo a more or less extensive review process.<\/p>\n

Constrained optimization<\/em> means that you not only want to minimize a target value (e.g. fuel consumption of a car), but also have to comply with one or more constraints (e.g. safety requirements for the structural integrity of the car). A while ago, Wolfgang Konen, together with Samineh Bagheri (former PhD student at TH K\u00f6ln) and with Prof. Thomas B\u00e4ck from Leiden University, developed the R package SACOBRA for constrained optimization, which is still one of the leading software packages in this field. A port of the R software to Python was carried out in 2025 just before June and it was successfully announced at the conference. It is now available for download from GitHub: https:\/\/github.com\/WolfgangKonen\/SACOBRA_Py<\/a>.<\/p>\n

\"\"<\/a>

The consistently mild climate produces trees with gigantic roots, nearly as large as Wolfgang Konen<\/p><\/div>\n

The MOPTA conference in the Azores offered also many opportunities for networking. Last but not least, the beautiful landscape of the Azores, was especially inviting.<\/p><\/div>\n

Kommentare deaktiviert f\u00fcr MOPTA (Modeling and Optimization) conference with invited speaker from TH K\u00f6ln<\/span><\/span>  |  Filed under: General<\/a>  |  Tags:<\/b> No Tag<\/div><\/div>

Invited speaker from TH K\u00f6ln at the MOPTA (Modeling and Optimization) conference<\/a><\/h2><\/div>
By Wolfgang Konen on Mai 19, 2025<\/div>
<\/p>\n
<\/div>\n

<\/a>The MOPTA conference (Modeling and Optimization: Theory and Applications, https:\/\/coral.ise.lehigh.edu\/mopta2025\/<\/a>) has been around for many years, since 2001 to be precise. It usually takes place in America, but this year, in 2025, European and American scientists can meet in the middle, so to speak, because it is being held in the Azores in June 2025.<\/p>\n

In June 2025, scientists from all over the world will meet on Sao Miguel, the main island of the Azores, to exchange scientific ideas. The research of the Gummersbach campus is attracting attention in this international circle, as Prof. Wolfgang Konen was invited by Don Jones, one of the co-organizers of the conference, to speak about his research area of constrained optimization as an invited speaker. This is a special honor, because normally one has to undergo a more or less extensive review process.<\/p>\n

Constrained optimization<\/em> means that you not only want to minimize a target value (e.g. fuel consumption of a car), but also have to comply with one or more constraints (e.g. concerning the statics of the car). Years ago, Wolfgang Konen, together with Samineh Bagheri (former PhD student) and Prof. Thomas B\u00e4ck from Leiden University, developed the R package SACOBRA for constrained optimization<\/em>, which is still one of the leading software packages in this field. A port of the R software to Python is currently in progress.<\/p>\n

The MOPTA conference in the Azores will also offer many opportunities for networking. Last but not least, the beautiful landscape of the Azores, which will also be visited on short excursions, will be inviting.<\/p>\n

\"\"<\/a>

Sete Cidades, the famous landscape formation (a green and a blue lake) in the Azores.<\/em><\/p><\/div><\/div>\n

Kommentare deaktiviert f\u00fcr Invited speaker from TH K\u00f6ln at the MOPTA (Modeling and Optimization) conference<\/span><\/span>  |  Filed under: Conference\/Journal paper<\/a>, MONREP<\/a>, Optimization<\/a>  |  Tags:<\/b> conference<\/a>, Constraint Optimization<\/a>, Deep Learning<\/a>, Optimization<\/a>, SACOBRA<\/a>, surrogates<\/a><\/div><\/div>

Iterative Decision Tree Learning in the Real World<\/a><\/h2><\/div>
By Raphael Engelhardt on Dezember 11, 2024<\/div>
<\/p>\n
<\/div>\n

<\/a>As mentioned in a previous blog post<\/a>, we developed an iterative algorithm for training decision trees (DTs) from trained deep reinforcement learning (DRL) agents. The algorithm combines the simple structure of DTs and the predictive power of well-performing DRL agents. In our publication<\/a>, we tested the idea on seven different control problems and successfully trained shallow DTs for each of these challenges, containing orders of magnitude fewer parameters than the DRL agents whose behavior they imitate.<\/p>\n

So far for the simulation... The real world is generally more challenging.<\/p>\n

Thanks to a fruitful collaboration with Prof. Tichelmann's Lab of Applied Artificial Intelligence, we were now able to put the idea to the test on a real-world robotics task. The lab offers a real-world implementation of the cart pole swing-up environment, a well-known benchmark for control problems and reinforcement learning. A physical pendulum is attached to a cart via an unactuated hinge. Only by swift movements of the cart to the left or to the right, the pendulum is first to be swung up and then balanced in the unstable equilibrium. During a previous bachelor's thesis, a DQN could be trained to solve the challenge successfully. We now used this DRL agent as oracle for our experiment. While the additional challenges of a real-world experiment were noticeable, the algorithm proved its robustness and managed to find a DT on par with the DQN agent, while using fewer parameters. Further details can be found in our latest paper<\/a>.<\/p>\n

A video shows the DT agent in operation.<\/p>\n

\nhttps:\/\/blogs.gm.fh-koeln.de\/ciop\/files\/2024\/12\/Iterative-DT-real-world-pendulum.webm<\/a><\/video><\/div><\/div>\n
Kommentare deaktiviert f\u00fcr Iterative Decision Tree Learning in the Real World<\/span><\/span>  |  Filed under: (RL)^3<\/a>, dataninja.nrw<\/a>, Publications<\/a>, Reinforcement Learning<\/a>  |  Tags:<\/b> AI<\/a>, dataninja.nrw<\/a>, Publications<\/a>, Reinforcement Learning<\/a>, Robotics<\/a><\/div><\/div>

Presenting our research at the World Conference on XAI<\/a><\/h2><\/div>
By Raphael Engelhardt on August 19, 2024<\/div>
<\/p>\n
<\/div>\n

<\/a>After having participated in its debut last year, it was a special pleasure to visit the second edition of The World Conference on Explainable Artificial Intelligence (xAI2024)<\/a>. The conference was a full immersion into all aspects of explainable AI. The keynote speech by Prof. Fosca Giannotti about hybrid decision-making and the two panel discussions on legal requirements of XAI and XAI in finance broadened the views between detailed poster and presentation sessions.<\/p>\n

On July 17th<\/sup>, I presented my work \"Exploring the Reliability of SHAP Values in Reinforcement Learning\"<\/a>, co-authored by Dataninja colleague<\/a> Moritz Lange and our supervisors Prof. Laurenz Wiskott and Prof. Wolfgang Konen.
\nThe work I presented is focused on using Shapley values for explainable reinforcement learning in multidimensional observation and action spaces, investigating questions about the reliability of approximation methods and the interpretation of feature importances. While Shapley values are a widely-used tool for machine learning, more work is required for its application to reinforcement learning. To those interested in Shapley values, I recommend to also take a look at the contribution of my Dataninja colleague Patrick Kolpaczki<\/a> on improving approximation of Shapley values. The conference proceedings are already available as part of Springer's book series \"Communications in Computer and Information Science\".<\/p>\n

Experiencing the conference at Valletta's (Malta) impressive Mediterranean Conference Center, learning about the work of newly met people, and reconnecting with familiar members of the XAI community from last year, has definitely been a highlight of this summer.<\/p>\n

\"\"<\/a>

View of capital Valletta on island Malta.<\/em><\/p><\/div>
\n

\"\"<\/a>

Guarded hallway to the seminar rooms.<\/em><\/p><\/div><\/p>\n

 <\/p><\/div>\n

Kommentare deaktiviert f\u00fcr Presenting our research at the World Conference on XAI<\/span><\/span>  |  Filed under: (RL)^3<\/a>, Conference\/Journal paper<\/a>, dataninja.nrw<\/a>, General<\/a>, Publications<\/a>  |  Tags:<\/b> (RL)^3<\/a>, AI<\/a>, conference<\/a>, dataninja.nrw<\/a>, Reinforcement Learning<\/a><\/div><\/div>

Dataninja Closing Conference<\/a><\/h2><\/div>
By Raphael Engelhardt on Juni 30, 2024<\/div>
<\/p>\n
<\/div>\n

<\/a>Time flies... It wasn't that long ago (or at least it feels like it) that I wrote a blog post about the first Dataninja Retreat<\/a>.<\/p>\n

Now we already held the closing conference<\/a> of the Dataninja project. From Tuesday 25th<\/sup> to Thursday 27th<\/sup> we had the pleasure to enjoy three days of science and meetups at Bielefeld University, the \u201cheadquarter\u201d of Dataninja. The rich program<\/a> consisted of keynote talks, poster sessions, and reports from our sibling project \u201cKI starters\u201d<\/a>.<\/p>\n

The (RL)3<\/sup> project<\/a> of Moritz Lange and supervisor Prof. Wiskott from Ruhr-University Bochum and myself under the supervision of Prof. Konen from TH K\u00f6ln, contributed with a short overview of our joint project and a more in-depth presentation of our most recent research in two poster contributions.<\/p>\n

Of special interest to our topics were the keynotes by Holger Hoos (\"How and Why AI will shape the future\"), Henning Wachsmuth (\"LLM-based Argument Quality Improvement\"), and Sebastian Trimpe (\"Trustworthy AI for Physical Machines\"). Many thanks to Prof. Barbara Hammer and her team (Dr. Ulrike Kuhl, \u00d6zlem Tan) from Bielefeld University for organizing and hosting such a fantastic event!<\/p>\n

As usual, it has been a very pleasant occasion to meet our fellow PhD candidates, and we have already made plans to meet up again, because the first ones are already on the home straight.
\n\"\"<\/a><\/div>\n

Kommentare deaktiviert f\u00fcr Dataninja Closing Conference<\/span><\/span>  |  Filed under: (RL)^3<\/a>, dataninja.nrw<\/a>, Reinforcement Learning<\/a>  |  Tags:<\/b> (RL)^3<\/a>, AI<\/a>, conference<\/a>, dataninja.nrw<\/a>, Reinforcement Learning<\/a><\/div><\/div><\/div><\/span><\/p>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":41,"featured_media":0,"parent":0,"menu_order":7,"comment_status":"open","ping_status":"open","template":"","meta":{"_acf_changed":false,"footnotes":""},"class_list":["post-630","page","type-page","status-publish","hentry"],"acf":[],"_links":{"self":[{"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/pages\/630","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/users\/41"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/comments?post=630"}],"version-history":[{"count":14,"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/pages\/630\/revisions"}],"predecessor-version":[{"id":644,"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/pages\/630\/revisions\/644"}],"wp:attachment":[{"href":"https:\/\/blogs.gm.fh-koeln.de\/ciop\/wp-json\/wp\/v2\/media?parent=630"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}