Fact-checked by Grok 2 weeks ago

Journal of Machine Learning Research

The Journal of Machine Learning Research (JMLR) is a peer-reviewed, open-access scientific journal dedicated to publishing high-quality scholarly articles across all areas of machine learning, serving as an international forum for both electronic and print dissemination of rigorous research in the field.^[1] Founded in 2000 by Leslie Pack Kaelbling as a nonprofit initiative to provide free access to machine learning literature, JMLR published its inaugural issue in March 2001 and quickly gained prominence, ranking first in computer science/artificial intelligence according to the 2002 ISI Journal Citation Reports.^[2] The journal pioneered open-access practices in academic publishing, making all papers freely available online immediately upon acceptance, with electronic versions under ISSN 1533-7928 and optional print volumes (ISSN 1532-4435) handled by Microtome Publishing since 2004.^[1] It emphasizes rapid yet thorough peer review, typically completing the process in about four months, and covers foundational topics such as statistical learning theory, deep learning, reinforcement learning, and applications in diverse domains.^[3] JMLR's editorial structure includes co-editors-in-chief Francis Bach (Inria) and David Blei (Columbia University), who have served since 2017, supported by over 90 action editors, specialized MLOSS editors for machine learning open-source software, and an advisory board of leading researchers like Geoffrey Hinton.^[4] Past editors-in-chief, including Kaelbling (2000–2009), Lawrence Saul (2009–2013), and Bernhard Schölkopf and Kevin Murphy (2013–2017), with Schölkopf serving until 2021, shaped its evolution.^[2] As a cornerstone of machine learning scholarship, JMLR boasts a high research impact, evidenced by an h-index of 280 and consistent recognition as a top-tier venue in the discipline, fostering advancements that have influenced fields from artificial intelligence to computational biology.^[5]

Introduction

Overview

The Journal of Machine Learning Research (JMLR) was established in 2000 as a non-profit, peer-reviewed academic journal dedicated to providing free electronic access to high-quality research in machine learning.^[1]^[2] It serves as an international forum for the publication of scholarly articles that advance the theory and methods of machine learning through principled contributions supported by empirical and theoretical validation.^[3] Founded by Leslie Pack Kaelbling, a computer science professor at MIT, JMLR has become a cornerstone venue for the global machine learning community.^[2] JMLR emphasizes rigorous scholarship that elucidates the underlying principles of machine learning systems and theoretical advancements, ensuring relevance to the field without focusing on domain-specific applications.^[3] Currently, the journal is led by Editors-in-Chief Francis Bach of Inria and David Blei of Columbia University, who oversee its commitment to excellence.^[6] With an electronic ISSN of 1533-7928 and a print ISSN of 1532-4435, it is published by the non-profit organization JMLR, Inc.^[1] At its core, JMLR operates on principles of accessibility and sustainability, offering immediate open access to all published papers without any author publication fees, thereby prioritizing the dissemination of knowledge over commercial interests.^[3]^[2] This model supports universal availability of research, fostering collaboration and innovation within the machine learning community.^[1]

Publication Details

The Journal of Machine Learning Research (JMLR) publishes all accepted articles electronically, making them freely available online immediately upon acceptance, with the journal's website launched on June 23, 2000, and the first papers published electronically starting in October 2000. Hosted on jmlr.org since its inception, the journal's website and content management migrated to GitHub in 2020 to enhance transparency and community involvement in maintenance.^[2]^[7] Print editions of JMLR were initially produced eight times per year by MIT Press from 2000 to 2004, with the first print issue released on March 30, 2001. Following the end of the MIT Press partnership, Microtome Publishing took over in 2004, shifting to an annual print-on-demand model for volumes, which continues to provide physical copies at a low cost without mandatory subscriptions.^[8]^[2] JMLR operates on a continuous publication model, organizing content into annual volumes starting from Volume 1 in 2000, with no imposed page limits or publication charges to authors. This structure supports rapid dissemination while maintaining archival organization.^[9] Regarding access, JMLR has been fully open access since its founding, with articles available under a permissive non-commercial reuse policy until 2017, when it adopted the Creative Commons Attribution (CC-BY) license to permit broader sharing and adaptation. This aligns with its open-access mission and has influenced related initiatives, such as the Proceedings of Machine Learning Research (PMLR) series, which originated as JMLR Workshop and Conference Proceedings in 2007.^[2]^[10]

History

Founding

The Journal of Machine Learning Research (JMLR) was founded in 2000 as a non-profit organization, JMLR, Inc., by Leslie Pack Kaelbling, a computer science professor at MIT, who served as the inaugural editor-in-chief, with David Cohn acting as managing editor.^[2] The primary motivations were to address the high subscription costs and publication delays associated with traditional journals, such as the Machine Learning journal published by Kluwer (now Springer), which charged institutions up to $1,006 annually and individuals $120, thereby restricting access to research in a rapidly expanding field.^[11]^[12] Inspired by the open-access model of the Journal of Artificial Intelligence Research (JAIR), JMLR aimed to provide free electronic access to high-quality peer-reviewed articles while promoting author retention of copyrights to enhance dissemination and reuse.^[2] The journal's website launched on June 23, 2000, enabling immediate online submissions and publications, followed by an official announcement on August 8, 2000, distributed via machine learning mailing lists, which outlined the scope, peer-review process, and commitment to rapid turnaround.^[2] The initial editorial structure included action editors such as Peter Bartlett from the Australian National University and Michael Jordan from UC Berkeley, alongside a broader international editorial board to ensure rigorous, expert review.^[2] This setup emphasized efficiency, with the goal of publishing accepted papers online promptly after review, contrasting the delays in commercial venues. On March 27, 2001, the Scholarly Publishing and Academic Resources Coalition (SPARC) endorsed JMLR as a model for researcher-led, affordable publishing, encouraging libraries to support it.^[11] The first print copies of Volume 1 arrived on March 30, 2001, produced in partnership with MIT Press, marking the journal's transition from digital-only to hybrid distribution while maintaining its core open-access commitment.^[2] Later in 2001, JMLR gained significant momentum when 40 members of the Machine Learning journal's editorial board, including prominent researchers like Geoffrey Hinton, Michael Jordan, and Tom Mitchell, resigned en masse on October 8, 2001, explicitly to support JMLR's open-access model and protest the barriers imposed by commercial publishing.^[12] This shift bolstered JMLR's credibility and board.

Key Milestones

In 2003, the Journal of Machine Learning Research achieved the top ranking (#1) in the ISI Journal Citation Reports for the Computer Science, Artificial Intelligence category, underscoring its rapid impact shortly after launch.^[2] The following year, in 2004, JMLR transitioned its print publication from MIT Press to Microtome Publishing after the agreement with MIT Press concluded, ensuring continued availability of physical copies. Additionally, a new review management software system, developed by Christian Shelton who had assumed the role of managing editor, was implemented to streamline the submission and peer-review processes; this system remains in use today.^[2] In 2007, JMLR launched the JMLR: Workshop and Conference Proceedings (W&CP) series to publish peer-reviewed papers from machine learning conferences and workshops, such as ICML and COLT, expanding its open-access offerings beyond the main journal.^[2] Lawrence Saul was appointed as Editor-in-Chief in 2009, guiding the journal through a period of sustained growth in submissions and influence.^[2] By 2013, Kevin Murphy and Bernhard Schölkopf were appointed as co-Editors-in-Chief, bringing expertise in probabilistic modeling and kernel methods to oversee editorial operations.^[2] In 2015, the W&CP series was renamed to Proceedings of Machine Learning Research (PMLR), aligning with broader plans to develop an umbrella of open-access publication outlets under the JMLR umbrella.^[2] The year 2017 marked several significant changes: JMLR adopted the Creative Commons Attribution (CC-BY) license for all new publications, enhancing reusability while maintaining open access; Kevin Murphy resigned as co-Editor-in-Chief in spring; Dave Blei joined as co-Editor-in-Chief in late 2017 to replace Murphy; and Francis Bach joined as co-Editor-in-Chief in early 2018.^[2]^[10] Finally, in December 2021, Bernhard Schölkopf retired as co-Editor-in-Chief after serving since 2013, concluding his contributions to the journal's editorial leadership.^[2] In 2023, Pradeep Ravikumar (Carnegie Mellon University) and Tong Zhang (University of Illinois Urbana-Champaign) were appointed as co-Editors-in-Chief.^[4]

Scope and Policies

Research Focus

The Journal of Machine Learning Research (JMLR) encompasses all areas of machine learning, with a primary emphasis on the development of new principled algorithms, theoretical foundations, empirical studies, and applications within artificial intelligence that advance core learning principles. Contributions typically include novel algorithms justified through empirical validation, theoretical, psychological, or biological insights; experimental or theoretical analyses providing new understanding of learning behaviors in intelligent systems; formalizations of emerging learning tasks in AI contexts; and analytical frameworks that enhance theoretical examinations of practical methods. This broad topical scope ensures the journal serves as an international forum for high-quality scholarly articles fostering foundational progress in the field.^[13]^[1] Submission criteria for JMLR require previously unpublished papers that demonstrate rigorous methodological novelty, such as sound empirical validation or theoretical analysis, while emphasizing reproducibility through clear, self-contained presentations of claims supported by experiments or derivations. The journal welcomes significant surveys only by editorial invitation, prioritizing work of interest to a wide machine learning audience over narrowly specialized topics. Authors must avoid simultaneous submissions to conferences or other venues, and papers are expected to contribute principled advancements rather than incremental improvements without substantial justification.^[13] JMLR explicitly excludes purely applied work lacking innovation in machine learning methodologies, insisting that all submissions must propel the computational or mathematical underpinnings of learning forward. Applications of machine learning techniques to non-machine-learning domains, such as specific industry uses without theoretical or methodological contributions, fall outside the journal's purview, as do unsolicited surveys or papers deemed too niche by the editorial board. This selective focus maintains the journal's commitment to impactful, broadly relevant research.^[13] JMLR has upheld a dedication to high-impact, archival-quality research in machine learning, evolving as an open-access platform that follows a 35-page guideline for submissions, with longer papers potentially facing rejection and those exceeding 50 pages requiring explicit justification to avoid desk rejection—and encourages supplementary materials like code to bolster reproducibility. This policy has supported the publication of comprehensive studies that might exceed conference constraints, reinforcing the journal's role in disseminating enduring contributions to the field.^[13]^[1]

Article Types

The Journal of Machine Learning Research (JMLR) primarily publishes full research articles in its main track, which focus on novel contributions to machine learning, including new algorithms, theoretical analyses, empirical studies, or new learning tasks that advance the field's understanding. These articles follow a 35-page guideline but longer submissions may be rejected, with those over 50 pages requiring explicit justification to avoid desk rejection.^[13] In addition to the main track, JMLR features special sections for targeted content. The Machine Learning Open Source Software (MLOSS) track supports the publication of peer-reviewed descriptions of high-quality open-source implementations of non-trivial machine learning algorithms, toolboxes, or related scientific computing tools, such as libraries for optimization or data processing; these contributions are now fully integrated as a dedicated section within the journal. Survey papers on emerging machine learning topics are also accepted, but only by invitation from the editorial board, ensuring they provide comprehensive overviews without accepting unsolicited submissions. Furthermore, JMLR occasionally organizes special issues on specific themes; these are managed like workshops with rolling submissions and were trialled starting in 2022, subject to review.^[14]^[13]^[15] JMLR integrates with conference proceedings through its sister series, the Proceedings of Machine Learning Research (PMLR), established in 2007 as JMLR Workshop and Conference Proceedings and rebranded to PMLR in 2017, to publish selected papers from major machine learning conferences like the International Conference on Machine Learning (ICML) and the International Conference on Artificial Intelligence and Statistics (AISTATS); this allows archival publication of conference content under the same open-access umbrella without duplicating JMLR's journal format. For extensions of prior conference papers submitted to JMLR's main track, substantial novelty—such as new results, deeper analysis, or broader insights—is required, with prior work clearly cited and differentiated.^[16]^[13]^[17] Other formats like technical reports or minor extensions are considered only if they offer significant novel contributions meeting the journal's standards for originality and impact; JMLR does not publish editorials, short communications, or review articles as primary types. All article types, including main track papers, MLOSS contributions, and invited surveys, undergo rigorous peer review by at least three reviewers assigned by an action editor, with decisions ranging from acceptance with minor revisions to rejection. Supplementary materials, such as code, datasets, and appendices, are strongly encouraged for all submissions to promote reproducibility, with source code releases requiring a formal agreement.^[13]^[18]^[19]

Editorial Structure

Editors-in-Chief

The Journal of Machine Learning Research (JMLR) was founded in 2000 under the leadership of its inaugural Editor-in-Chief, Leslie Pack Kaelbling, who served from 2000 to 2009. Kaelbling, a professor at MIT, played a pivotal role in establishing JMLR as a pioneering open-access journal, emphasizing free electronic distribution and rapid peer review to democratize access to machine learning research. Her tenure focused on building the journal's foundational infrastructure, including partnerships with publishers like MIT Press and the adoption of a 6-week review turnaround, which set JMLR apart from traditional subscription-based outlets.^[2] Lawrence Saul succeeded Kaelbling as Editor-in-Chief, serving from 2009 to 2013. During his term at the University of California, San Diego, Saul oversaw significant growth in manuscript submissions and the journal's rising prominence in academic rankings, solidifying JMLR's reputation as a leading venue in the field. This period marked an expansion in the journal's scope and operational scale, aligning with the burgeoning interest in machine learning across disciplines.^[2]^[20] From 2013 to spring 2017, JMLR adopted a co-Editor-in-Chief model with Kevin Murphy (University of British Columbia) and Bernhard Schölkopf (Max Planck Institute for Intelligent Systems), emphasizing a balance between theoretical advancements and applied methodologies. Murphy's term ended in spring 2017, while Schölkopf continued, joined by David Blei in late 2017 until Schölkopf's retirement in December 2021; together, they introduced innovations such as alternating paper assignments for efficiency and special tracks for conference proceedings, like the 2013 ICLR issue, to bridge rigorous theory with practical impacts.^[2] The current co-Editors-in-Chief, Francis Bach from Inria (appointed in early 2018) and David Blei from Columbia University (appointed in late 2017), have guided JMLR since Bach's appointment, with Blei and Bach leading together following Schölkopf's retirement in December 2021. Bach specializes in optimization and statistical machine learning, and Blei in probabilistic modeling and topic models. Their leadership has maintained the journal's commitment to open access, including a 2020 website migration to GitHub for enhanced accessibility and sustainability.^[2]^[21] Editors-in-Chief are appointed through a process involving recommendations from prior leaders or search committees formed by the editorial board, typically for terms of 3 to 5 years, though extensions occur based on continuity needs. Their primary responsibilities include setting editorial policies, overseeing the review process at a high level, and ensuring the journal's strategic direction aligns with evolving standards in machine learning research.^[2]

Review Process

The Journal of Machine Learning Research (JMLR) employs a double-blind peer review process to ensure impartial evaluation, where submissions are anonymized and author identities are concealed from reviewers and action editors throughout the review cycle.^[13] Upon submission, the Editor-in-Chief conducts an initial scan to check for basic standards and scope alignment before assigning the manuscript to an action editor, who oversees the detailed review.^[13] The action editor then selects typically three external reviewers with relevant expertise to provide in-depth assessments.^[13] The review process unfolds in several stages: following assignment, the action editor may issue an early rejection with a brief justification if the paper is deemed clearly unsuitable; otherwise, it proceeds to full external review.^[13] Reviewers evaluate the submission based on criteria including the clarity and achievability of research goals, the replicability of methods and results, adequate theoretical or empirical validation, overall significance and novelty in advancing machine learning, thorough discussion of related work with acknowledgment of limitations, and general readability with precise explanations of technical elements.^[19] Upon receipt of reviews, the action editor renders the final decision—accept (possibly with minor revisions), reject, or invite resubmission after substantial revisions—with no formal author rebuttal phase.^[13] JMLR commits to rigorous yet thorough reviewing, with first-round reviews typically taking 3-4 months, though timelines can vary for complex papers or high-volume periods, with overall decisions often taking several months to a year.^[22] Key policies include no submission or publication fees, making the process accessible without financial barriers, and a strong emphasis on reproducibility, where authors are encouraged (but not required) to provide open-source code or data to support empirical claims, often through JMLR's dedicated MLOSS track for software contributions.^[13]^[18] Submissions must adhere to strict originality rules, prohibiting simultaneous submissions elsewhere and mandating disclosure of prior related work or funding sources that could pose conflicts.^[13] The journal utilizes a custom electronic submission and review management system, developed in 2004 and maintained in-house, which handles the workflow for hundreds of annual submissions while publishing over 250 papers per year.^[22]^[20]

Impact and Recognition

Citation Metrics

The Journal of Machine Learning Research (JMLR) has an impact factor of 5.2, as reported in the 2025 Journal Citation Reports by Clarivate Analytics. This metric reflects the average number of citations received by articles published in the journal over the preceding two years, underscoring its steady high ranking within artificial intelligence and machine learning categories. Previously, the impact factor stood at 4.091 in 2018, demonstrating consistent growth and influence in the field.^[23] JMLR's h-index is 280 as of 2025, signifying that 280 articles from the journal have each garnered at least 280 citations.^[5] This robust h-index highlights the enduring impact of its publications across diverse machine learning subdomains. Additional metrics further affirm JMLR's prominence, including an SCImago Journal Rank (SJR) of 2.019 for 2024, which positions it in the Q1 quartile for both software and artificial intelligence categories based on Scopus data.^[5] Its CiteScore is 16.3 (2024), measuring the average citations per document over a four-year window.^[24] Overall, the journal has accumulated more than 591,000 total citations since its inception (Scopus, as of 2024).^[25] JMLR achieved the top ranking in its category in 2002 and has maintained top-tier status through 2025, bolstered by its open access model that enhances global accessibility and citation potential.^[5]^[1]

Influence on Field

The Journal of Machine Learning Research (JMLR) pioneered open access publishing in the machine learning field by making all articles freely available online since its inception in 2000, thereby democratizing access to high-quality research and reducing barriers for researchers worldwide, particularly in resource-limited regions.^[1] This model influenced the broader academic publishing landscape, inspiring the creation of the Proceedings of Machine Learning Research (PMLR), which originated as JMLR's workshop and conference proceedings series and adopted similar open access principles to archive conference outputs.^[16] Similarly, the Transactions on Machine Learning Research (TMLR), launched in 2022 as a complementary venue to JMLR, extends this commitment by emphasizing rapid, open review processes and electronic dissemination to support emerging ML topics.^[26] JMLR has hosted numerous seminal publications that advanced core areas of machine learning. In the early 2000s, it published influential extensions to support vector machines, such as active learning approaches that optimized label acquisition for SVMs, enhancing efficiency in large-scale classification tasks.^[27] During the 2010s, foundational deep learning works appeared, including the introduction of dropout as a regularization technique to prevent overfitting in neural networks, which became a standard method in training deep architectures.^[28] High-impact contributions in probabilistic models, like the neural probabilistic language model that laid groundwork for modern word embeddings, and in reinforcement learning, such as analyses of multi-agent settings, further solidified JMLR's role in shaping theoretical and practical advancements.^[29]^[30] Within the ML community, JMLR functions as a prestigious archival venue, often serving as the long-term home for expanded versions of papers initially presented at conferences like ICML and COLT, ensuring enduring accessibility and citation.^[31] Its Machine Learning Open Source Software (MLOSS) track promotes reproducibility by publishing vetted implementations of algorithms and toolboxes, encouraging the open-source ethos that underpins reliable ML research.^[14] JMLR papers have frequently received recognition through best paper awards at major conferences, reflecting their quality and influence on subsequent work.^[32] This legacy extends to education and industry, where JMLR contributions inform ML curricula at universities and underpin tools adopted in real-world applications, from probabilistic inference systems to scalable learning frameworks. As of 2025, JMLR continues to propel AI advancements, with its publications integrated into journal-to-conference tracks at NeurIPS, ICML, and ICLR, allowing archival JMLR works to gain visibility through conference presentations and thereby influencing evolving standards in ML research dissemination.^[31]

References

[1]
Journal of Machine Learning Research
The Journal of Machine Learning Research (JMLR), established in 2000, provides an international forum for the electronic and paper publication of high-quality ...JMLR PapersJMLR Information for Authors
[2]
History of JMLR
Announcing the JOURNAL of MACHINE LEARNING RESEARCH · Editor: Leslie Pack Kaelbling · Managing Editor: David Cohn · Action Editors: · Peter Bartlett, Australian ...
[3]
https://www.jmlr.org/author-info.html
[4]
JMLR Editorial Board
JMLR Action Editors. Edo Airoldi, Harvard University, USA Statistics, approximate inference, causal inference, network data analysis, computational biology ...Missing: impact | Show results with:impact
[5]
Journal of Machine Learning Research - Scimago
Scope. The Journal of Machine Learning Research (JMLR) provides an international forum for the electronic and paper publication of high-quality scholarly ...
[6]
About: Journal of Machine Learning Research
It was established in 2000 and the first editor-in-chief was Leslie Kaelbling. The current editors-in-chief are Francis Bach (Inria) and David Blei (Columbia ...
[7]
https://github.com/JmlrOrg/webpage
[8]
Journal of Machine Learning Research and Microtome Publishing ...
Oct 12, 2004 · ... MIT Press, providing eight issues a year at a $400 annual ... The open access Journal of Machine Learning Research was founded by ...
[9]
https://www.jmlr.org/papers/
[10]
https://jmlr.org/forms/jmlr-license-agreement-2017.pdf
[11]
Press Release - Journal of Machine Learning Research
Mar 27, 2001 · A number of Machine Learning editorial board members have resigned to join the editorial board of JMLR. SPARC, an alliance of libraries that ...<|control11|><|separator|>
[12]
Leading ML researchers issue statement of support for JMLR
... journal Dear colleagues in machine learning, The forty people whose names appear below have resigned from the Editorial Board of the Machine Learning Journal ( ...Missing: founding | Show results with:founding
[13]
JMLR Information for Authors
When a paper is submitted to JMLR, it is scanned by the Editor-in-Chief (EIC). If the EIC finds that the paper is very clearly below the standards of the ...
[14]
Machine Learning Open Source Software
Machine Learning Open Source Software ... To support the open source software movement, JMLR MLOSS publishes contributions related to implementations of non- ...Open Source Software
[15]
Journal of Machine Learning Research Special Issues
JMLR is trialling a new procedure for Special Issues. This procedure will be in place for 1 year (2022), after which time it will be reviewed.Missing: types sections surveys
[16]
Proceedings of Machine Learning Research | The Proceedings of ...
The Proceedings of Machine Learning Research is a series that publishes machine learning research papers presented at conferences and workshops.
[17]
PMLR - Proceedings of Machine Learning Research
The Proceedings of Machine Learning Research (formerly JMLR Workshop and Conference Proceedings) is a series aimed specifically at publishing machine ...
[18]
Machine Learning Open Source Software
JMLR is proud to support the open source movement by having a track on open source software in machine learning.
[19]
Guidelines for JMLR reviewers
Guidelines for JMLR reviewers. Please touch upon as many of the following points as practical: Goals: What are research goals and learning task?Missing: types sections surveys
[20]
Retrospectives from 20 Years of JMLR
Feb 21, 2022 · In 2000, led by editor-in-chief Leslie Kaelbling, JMLR was founded as a fully free and open-access platform for publishing high-quality machine ...
[21]
JMLR Editorial Board
David Abel, reinforcement learning, philosophy, planning, abstraction; Evrim Acar, matrix/tensor factorizations; Maximilian Alber, deep learning, semantic ...
[22]
https://jmlr.org/history.html
[23]
Journal of Machine Learning Research Impact Factor IF 2025 - Bioxbio
Journal of Machine Learning Research Impact Factor, IF, number of article, detailed information and journal factor ... 2023, 4.3, -, -. 2022, 6.0, -, 50577. 2021 ...
[24]
Transactions on Machine Learning Research
Editors-in-Chief of TMLR are Hugo Larochelle (Mila), Naila Murray (Meta), Gautam Kamath (University of Waterloo), and Nihar B. Shah (CMU). Founding Editors-in- ...Editorial Board · Accepted papers · TMLR Submission Instructions · FAQs
[25]
[PDF] Support Vector Machine Active Learning with Applications to Text ...
This paper introduces a new algorithm for active learning with SVMs, where the learner chooses which unlabeled instances to request labels from, reducing the ...Missing: seminal | Show results with:seminal
[26]
https://jmlr.org/tmlr/
[27]
[PDF] A Neural Probabilistic Language Model
Abstract. A goal of statistical language modeling is to learn the joint probability function of sequences of words in a language.
[28]
[PDF] Multi-Agent Reinforcement Learning in Common Interest and Fixed ...
Abstract. Multi Agent Reinforcement Learning (MARL) has received continually growing attention in the past decade. Many algorithms that vary in their ...
[29]
The NeurIPS/ICLR/ICML Journal-to-Conference Track
The track considers certain published papers from the Journal of Machine Learning Research (JMLR) and the Transactions on Machine Learning Research (TMLR), for ...
[30]
Submission Guidelines and Editorial Policies
TMLR publishes original papers on learning principles, new algorithms, and studies of learning in intelligent systems. It does not accept expanded conference ...