JMLR

JMLR
Title	Journal of Machine Learning Research
Abbreviation	JMLR
Discipline	Machine learning
Publisher	Microtome Press
Country	United States
History	2000–present
Frequency	Continuous
Issn	1532-4435

JMLR is an open-access, peer-reviewed scientific journal specializing in machine learning research and related areas. It publishes original articles, surveys, and technical reports that contribute to theory, algorithms, and applications across artificial intelligence and data-driven fields. The journal emphasizes rigorous methods, reproducible results, and broad dissemination to researchers working in statistics, computer science, robotics, and cognitive science.

History

JMLR was founded in 2000 amid debates involving NeurIPS, ICML, COLT, AAAI, and publishing practices associated with Springer, Elsevier, Kluwer Academic Publishers, MIT Press, and IEEE. The founding movement included researchers affiliated with Stanford University, Massachusetts Institute of Technology, Carnegie Mellon University, University of Toronto, University of California, Berkeley, University of Washington, Princeton University, Harvard University, Yale University, Caltech, Oxford University, Cambridge University, ETH Zurich, University College London, University of Edinburgh, Ecole Normale Supérieure, École Polytechnique Fédérale de Lausanne, Max Planck Society, CNRS, and INRIA. Early editorial leadership featured faculty with ties to Andrew Ng, Yoshua Bengio, Geoffrey Hinton, Michael I. Jordan, and David MacKay movements, reacting to subscription models used by Springer Science+Business Media and Wiley. Over time, JMLR established relationships with conferences such as NeurIPS 2000, ICML 2001, AISTATS, UAI, and KDD, while receiving submissions from authors at labs like Google Brain, DeepMind, OpenAI, Facebook AI Research, Microsoft Research, IBM Research, and Amazon Web Services.

Scope and content

The journal covers topics spanning supervised learning, unsupervised learning, reinforcement learning, probabilistic modeling, optimization, kernel methods, ensemble methods, deep learning, representation learning, Bayesian inference, causal inference, online learning, and structured prediction. Authors often reference work from researchers at Bell Labs, AT&T Research, Siemens, NVIDIA, Intel Labs, Adobe Research, Sony CSL, Baidu Research, Tencent AI Lab, Alibaba DAMO Academy, and academic groups at Brown University, Duke University, University of Michigan, Columbia University, New York University, Cornell University, Rutgers University, Northwestern University, Rice University, University of Illinois Urbana–Champaign, Peking University, Tsinghua University, Seoul National University, KAIST, University of Tokyo, University of Melbourne, University of Sydney, Monash University, Indian Institute of Technology Bombay, Indian Institute of Science, and National University of Singapore.

Publication and access model

JMLR operates a continuous publication model with open-access policies that contrast with paywalled venues like Nature, Science, Cell, The Lancet, and some journals of Springer Nature. The publisher entity known as Microtome Press and the editorial board manage article processing without charging author-side fees in many cases, positioning the journal alongside initiatives such as arXiv, HAL, bioRxiv, and institutional repositories at Stanford Libraries and MIT Libraries. The journal's infrastructure interoperates with indexing services and citation aggregators including Google Scholar, Scopus, Web of Science, and Semantic Scholar.

Editorial board and peer review

The editorial board comprises scholars and researchers drawn from institutions like Stanford University, MIT, Carnegie Mellon University, University of Toronto, ETH Zurich, University College London, Oxford University, Cambridge University, Princeton University, Harvard University, Yale University, University of California, Berkeley, University of Washington, Peking University, and Tsinghua University. The peer-review process is single- or double-blind depending on submission type and involves reviewers who are active participants in conferences such as NeurIPS, ICML, AISTATS, KDD, ACL, CVPR, ICLR, ECCV, and SIGIR. Editorial practices reflect standards discussed at meetings of organizations including ACM, IEEE, AAAI, SIAM, IMS, and ISBA.

Impact and reception

JMLR is widely cited in literature spanning theoretical computer science, applied statistics, robotics, natural language processing, computer vision, bioinformatics, and neuroscience. The journal's articles influence work at industrial labs such as Google DeepMind, OpenAI, Facebook AI Research, Microsoft Research Redmond, IBM Watson, and Apple Machine Learning Research, and are often taught in courses at MIT, Stanford, UC Berkeley, Carnegie Mellon, Oxford, Cambridge, ETH Zurich, University of Toronto, and National University of Singapore. Citation metrics and impact assessments reference databases maintained by Clarivate Analytics, Eigenfactor Project, Scimago, and Google Scholar Metrics.

Notable papers and contributions

JMLR has published influential papers on support vector machines, boosting algorithms, kernel methods, Bayesian nonparametrics, Gaussian processes, graphical models, Markov chain Monte Carlo, stochastic gradient methods, and deep learning theory. Landmark contributions have been connected to researchers associated with Vladimir Vapnik, Corinna Cortes, Leo Breiman, Yoav Freund, Robert Schapire, Michael Jordan, Radford Neal, Zoubin Ghahramani, Christopher Bishop, David MacKay, Tomaso Poggio, Yann LeCun, Geoffrey Hinton, Yoshua Bengio, Ian Goodfellow, Sergey Levine, Pieter Abbeel, Daphne Koller, Andrew Ng, Trevor Hastie, Rob Tibshirani, Jerome Friedman, Shai Shalev-Shwartz, Nathan Srebro, Ben Taskar, John Lafferty, Tom Mitchell, Peter Dayan, Marcus Hutter, Jurgen Schmidhuber, Sepp Hochreiter, Christopher Manning, Fei-Fei Li, Jitendra Malik, Antonio Torralba, Aude Oliva, Gary B. Huang, Li Fei-Fei, William Cohen, Noah Goodman, Pietro Perona.

Category:Academic journals