Slides and videos of the last three talks of the JC are now out 🔥🔥🔥: 👉 sml.disi.unitn.it/journal.html

andrea passerini introducing the IJCAIconf crowd to probabilisitic inference under algebraic and logical constraints Paolo Morettin Samuel Kolb Structured Machine Learning Group - Trento DTAI AASS MPI Lab (dtai.cs.kuleuven.be/tutorials/wmit…)

📢 Our work on handling #drift in #hierarchical classification through #interactive #ML is now available on Data Mining and Knowledge Discovery! link.springer.com/article/10.100… Joint work with Andrea Bontempelli (Andrea), Fausto Giunchiglia (Knowdive Group), and Andrea Passerini (andrea passerini).

I am happy to share that our paper, GlanceNets, has been accepted at #NeurIPS22!! It is my first work under the supervision of andrea passerini and Stefano Teso. More information about it coming soon 🙃

🔔[weekend reading]🔔 Really happy to announce the first Global Explainer for GNN, capable of generating explanations as arbitrary Boolean combinations of learned graphical concepts. arxiv.org/abs/2210.07147 Steve Azzolin Pietro Barbiero Pietro Lio' andrea passerini

After months of experiments we are happy to announce our last work: "Explaining the explainers in GNN: a comparative study" arxiv.org/abs/2210.15304 Steve Azzolin Gabriele Santin @_giuliacencetti @pl219_Cambriin Bruno Lepri andrea passerini 1/4

Can we learn Deep Neural Networks that explain their own predictions using human-interpretable #concepts? Yes! Just read our new #NeurIPS2022 paper: 𝙂𝙡𝙖𝙣𝙘𝙚𝙉𝙚𝙩𝙨: Interpretable, Leak-proof Concept-based model Stefano Teso andrea passerini 🧵1/9

I am thrilled to announce that both of our #ICLR2023 submissions have been accepted, with one being selected as a notable-top-25%! Congratulations to all co-authors and stay tuned for further details. Structured Machine Learning Group - Trento

Our work about explaining GNNs via Logic has been accepted at #ICLR2023 🎉 While waiting for the conference, you can check out my talk here youtu.be/7dCrlFDqeB4 arxiv.org/abs/2210.07147 Antonio Longa Pietro Barbiero Pietro Lio' andrea passerini

GNN-based models for temporal networks. SOTA, a taxonomy, open challenges and opportunities. Here is the link arxiv.org/abs/2302.01018 With Veronica Lachi, Gabriele Santin, Monica Bianchini, Bruno Lepri, Pietro Lio', Franco Scarselli, andrea passerini

🚨Our paper "Synthesizing explainable counterfactual policies for algorithmic recourse with program synthesis" is now available on Machine Learning (rdcu.be/c4Z0V)! We show how to generate interventions to overturn predictions of ML models to benefit the affected users!

Super excited to announce that our #TANGO Horizon Europe🇪🇺 RIA proposal was selected for funding (with a score of 15/15!!) Looking forward to working on hybrid human-machine learning and decision making with a stellar consortium!

Following our paper on algorithmic recourse (rdcu.be/c72Gb), we just released an updated version of the code of the experiments to improve the usability of our FARE methods (github.com/unitn-sml/reco…)! Structured Machine Learning Group - Trento UniTrento_DISI Fondazione Bruno Kessler - FBK ELLIS (1/2)

Such an amazing presentation by Irene Cannistraci and Luca Moschella at AI Trento Journal Club. "Communicating between latent spaces with limited semantic correspondence" Structured Machine Learning Group - Trento Emanuele Marconato Giovanni De Toni Nicola Dall'Asen Gianluca Apriceno @FranciF1997 Steve Azzolin

Thrilled that our workshop proposal of “Hybrid Human-Machine Learning and Decision Making (HLDM)” has been accepted to be co-located with ECML PKDD ! 🎉 Looking forward to organising it with andrea passerini , fabio casati , Anna Monreale , Roberto Pellungrini, and Paula Gürtler!

Can we leverage prior knowledge to improve Continual Learning? Yes, with NeSy-CL! I'm happy to introduce our paper #NeuroSymbolic #ContinualLearning, to appear at #ICML2023 🌴 A work with gianpaolo bontempo, simone calderara, Elisa Ficarra, andrea passerini, and Stefano Teso 🧵1/n

📢 Exciting news, folks! 🎉 Registration is now open for the Trento Learning on Graphs Conference 2025 Meetup happening from November 27th to 30th, 2023! 📝 Secure your spot here 👉 bit.ly/TrentoLogMeetu… For all the call details and more, visit our event website 👉 bit.ly/TrentoLogMeetu…

How can we better personalize Algorithmic Recourse recommendations? ✨ In our latest TMLR (Transactions on Machine Learning Research) paper, we show PEAR, the first human-in-the-loop approach for customizing recourse suggestions to any user. 👉openreview.net/forum?id=8sg2I… Stefano Teso andrea passerini Bruno Lepri

📣 Call for Papers: The 2nd Workshop on Hybrid Human-Machine Learning & Decision Making (HLDM’24). 🗓️ Deadline: June 15, 2024. 🌟 Confirmed Keynotes: Prof. Nick Chater, Prof. Zeynep Akata, and Prof. nuriaoliver 🔗 Website: sml.disi.unitn.it/hldm24.html

Just landed in Athens 🇬🇷 for #FAccT25! I’ll be presenting our work on the fragility of causal recourse over time (arxiv.org/pdf/2410.08007) on Monday morning! If you are working on algorithmic recourse, explainability, and/or responsible AI, let’s grab a ☕ at the conference!