Mehmet Tunçel

Öğretim Üyemizin Yer Aldığı Uluslararası Projeye Ufuk Avrupa MSCA-SE Fon Desteği!🔬 Yapay Zeka ve Veri Bilimi Uygulama ve Araştırma Merkezi (İTÜ AI) öğretim üyemiz Dr. Mehmet Tunçel’in ortağı olduğu “ExCISC” projesi, Avrupa Komisyonu Horizon Europe MSCA-SE programı kapsamında desteklenmeye hak kazandı. Dublin Teknoloji Üniversitesi koordinatörlüğünde 48 ay sürecek proje; emniyet açısından kritik alanlarda insan kararlarını destekleyen, etik ve insan merkezli "iş birlikçi zekâ" sistemlerinin geliştirilmesini hedefliyor. 8 ülkeden 20 ortağın yer aldığı uluslararası konsorsiyum ile akademi ve sanayi arasında teknoloji transferi ve ortak Ar-Ge kapasitesinin artırılması amaçlanıyor.🎯 Öğretim üyemiz Dr. Mehmet Tunçel'i tebrik ediyor, başarılarının devamını diliyoruz. #ITUGURUR 👏

I am happy to announce that the first draft of my RL tutorial is now available. arxiv.org/abs/2412.05265

(1/7) Topological machine learning is a fascinating field that promises to harness the higher-order interactions encountered in complex systems, enabling new horizons to be explored and older ones to be revisited. However, in order to drive this progress, there is a critical need for topological benchmark datasets.

If this doesn’t make you sick, I don’t know what to say to you…

Make up for the shameful decision of USA, Canada and others and donate to @UNRWA at donate.unrwa.org/-landing-page/…

Prof. Gil Strang's last lecture is ... Monday! Watch it via live stream: grinfeld.org/strang/

🎉A unique masterpiece book which combines information theory and machine learning: "Information Theory, Inference, and Learning Algorithms" by Sir MacKay 🆓 PDF inference.org.uk/itila/book.html with 16 video lectures 🎥: youtube.com/playlist?list=… Worth checking out!

Designing, Visualizing, and Understanding Deep Neural Networks. UC Berkeley. 100% Free. 66 YouTube videos: youtube.com/playlist?list=…

Videos from @nla_grp New Directions in Numerical Linear Algebra and High Performance Computing Workshop in honour of Jack Dongarra nhigham.com/2021/08/17/vid…

Unsupervised Learning for Predictive Maintenance using Auto-Encoders - KDnuggets buff.ly/2N6DdF1

👌

Lloyd Trefethen has put his excellent book on Approximation Theory online in pdf form at tinyurl.com/y3o2o26b

Using L-BFGS to solve problems with millions of variables? NAG's First Order Active Set (FOAS) solver is a new method for large-scale unconstrained or bound-constrained nonlinear optimization problems. FOAS beats L-BFGS in memory requirements and speed nag.com/content/limite…

Short explanations of numerical linear algebra terms by prof. @nhigham github.com/higham/what-is

NVC++ is our new compiler. It can accelerate your Standard C++ code with GPUs. Keep your eye on this. We're Executing on big and exciting plans for it. developer.nvidia.com/blog/accelerat… @NVIDIAHPCDev

Thrilled to announce another global addition to the DeepMind scholarship programme with @itu_ai_research! Read more about the qualifications, benefits, application process and timeline today: bit.ly/3iUYkEL

It looks like Springer is offering a lot of textbooks for free during the pandemic (the electronic versions, at least). I was happy to see that one of my favorites "Proofs from the Book" is among them. link.springer.com/search?facet-c…

Cool work. Static analysis in GCC 10 developers.redhat.com/blog/2020/03/2…

SIR models are cool, but I don't understand why they should be modeled with an ODE. A Markov chain can represent the discreteness of reality, and makes the likelihood faster to compute than solving an ODE at each iteration. Are there reasons to prefer an ODE for this?