May's here and so are new #MLJ online-first papers: "Faster Riemannian Newton-type optimization by subsampling and cubic regularization" by Yian Deng & Tingting Mu (https://link.springer.com/article/10.1007/s10994-023-06321-0) (OA)

An #MLJ online-first #NewPaper on a new data set: "ROAD-R: the autonomous driving dataset with logical requirements" by Eleonora Giunchiglia, Mihaela Cătălina Stoian, Salman Khan, Fabio Cuzzolin & Thomas Lukasiewicz (https://link.springer.com/article/10.1007/s10994-023-06322-z)

#MLJ online-first #NewPaper: "Data driven discovery of systems of ordinary differential equations using nonconvex multitask learning" by Clément Lejeune, Josiane Mothe, Adil Soubki & Olivier Teste (https://rdcu.be/daInh)

Two #MLJ online-first #NewPaper|s on understanding models for image classification today: "Understanding CNN fragility when learning with imbalanced data" by Damien Dablain, Kristen N. Jacobson, Colin Bellinger, Mark Roberts & Nitesh V. Chawla (https://link.springer.com/article/10.1007/s10994-023-06326-9) (OA)

Good Friday #MLJ online-first paper: "An accelerated proximal algorithm for regularized nonconvex and nonsmooth bi-level optimization" by Ziyi Chen, Bhavya Kailkhura & Yi Zhou (https://rdcu.be/c9tDx)

New #MLJ online-first paper: "Robust matrix estimations meet Frank–Wolfe algorithm" by Naimin Jing, Ethan X. Fang & Cheng Yong Tang (https://rdcu.be/c9mY5)

Another #MLJ online-first #NewPaper dropped yesterday: "Domain adversarial neural networks for domain generalization: when it works and how to improve" by Anthony Sicilia, Xingchen Zhao & Seong Jae Hwang (https://link.springer.com/article/10.1007/s10994-023-06324-x) (OA)

We've a new #MLJ online-first paper: "Imbalanced gradients: a subtle cause of overestimated adversarial robustness" by Xingjun Ma, Linxi Jiang, Hanxun Huang, Zejia Weng, James Bailey & Yu-Gang Jiang (https://rdcu.be/c8P3h)

You thought we were done with #MLJ online-first #NewPaper|s this week? Well, we're not: "PreCoF: counterfactual explanations for fairness" by Sofie Goethals, David Martens & Toon Calders (https://rdcu.be/c8Fhc)

Do you like #ML papers? Because we at #MLJ have some for you: "Generalizing universal adversarial perturbations for deep neural networks" by Yanghao Zhang, Wenjie Ruan, Fu Wang & Xiaowei Huang (https://rdcu.be/c8A0j)

Very interesting #MLJ online-first paper on, essentially, composing music: "Deep learning’s shallow gains: a comparative evaluation of algorithms for automatic music generation" by Zongyu Yin, Federico Reuben, Susan Stepney & Tom Collins (https://link.springer.com/article/10.1007/s10994-023-06309-w) (OA)

A somewhat adversarial #MLJ online-first #NewPaper today: "Learning key steps to attack deep reinforcement learning agents" by Chien-Min Yu, Ming-Hsin Chen & Hsuan-Tien Lin (https://rdcu.be/c76Lb)

Isn't it annoying when the authors beat you to tweeting out their own #MLJ paper?! 🤨
---
RT @kerstingAIML
🚨New MLJ @MLJ_Social paper🚨 True, ChatGPT & GPT4 are impressive, still there is a gap to be filled on complex visual reasoning & learning. #alphaILP helps to close it: object-centric representations + differentiable forward chaining+structure learning
👉https://link.springer.com/article/10.1007/s10994-023-06320-1
https://twitter.com/kerstingAIML/status/1636056886045814786

Let's talk about #MLJ online-first #NewPaper|s, or, rather, tweet about them: "Partial counterfactual identification and uplift modeling: theoretical results and real-world assessment" by Théo Verhelst, Denis Mercier, Jeevan Shrestha & Gianluca Bontempi (https://rdcu.be/c7YTX)

New #MLJ online-first paper just dropped: "SETAR-Tree: a novel and accurate tree algorithm for global time series forecasting" by @rakshithawg, @giwebb, Daniel Schmidt & @CBergmeir (https://link.springer.com/article/10.1007/s10994-023-06316-x) (OA)

I hope you like #MLJ online-first #NewPaper|s because we've more for you: "UnbiasedNets: a dataset diversification framework for robustness bias alleviation in neural networks" by Mahum Naseer, Bharath Srinivas Prabakaran, Osman Hasan & Muhammad Shafique (https://link.springer.com/article/10.1007/s10994-023-06314-z)

Another #MLJ online-first #NewPaper published just after I tweeted last: "Early anomaly detection in time series: a hierarchical approach for predicting critical health episodes" by Vitor Cerqueira, @Luis Torgo & @csoares (https://rdcu.be/c6DaG)

End-of-February online-first #MLJ #NewPaper|s: "Inverse learning in Hilbert scales" by Abhishake Rastogi & Peter Mathé (https://link.springer.com/article/10.1007/s10994-022-06284-8) (OA)

Quick #MLJ online-first #NewPaper update: "On the sample complexity of actor-critic method for reinforcement learning with function approximation" by Harshat Kumar, Alec Koppel & Alejandro Ribeiro (https://rdcu.be/c5RA4)

The days stay cold but you can huddle up with an #MLJ online-first #NewPaper: "PAC-learning with approximate predictors" by Andrew J. Turner & Ata Kabán (https://link.springer.com/article/10.1007/s10994-023-06301-4) (OA)