International Conference on Machine Learning

International Conference on Machine Learning
Name	International Conference on Machine Learning
Abbreviation	ICML
Discipline	Machine learning
Publisher	Proceedings of Machine Learning Research
First	1980
Frequency	Annual

Contents

International Conference on Machine Learning The International Conference on Machine Learning is a leading annual academic conference for researchers in Artificial intelligence, Machine learning (field), Statistics, Computer science, and related areas. It attracts attendees from institutions such as Google, Microsoft Research, OpenAI, DeepMind, Facebook AI Research, and universities including Massachusetts Institute of Technology, Stanford University, University of California, Berkeley, Carnegie Mellon University, and University of Toronto. ICML has shaped developments alongside venues like NeurIPS, AAAI Conference on Artificial Intelligence, International Joint Conference on Artificial Intelligence, KDD, and COLT.

History

ICML traces roots to early workshops and symposia in the late 1970s and 1980s, following gatherings such as the Conference on Automated Learning and Discovery and meetings held by Association for Computing Machinery special interest groups. Early contributors included researchers from Bell Labs, AT&T Labs, IBM Research, SRI International, and universities like University of Edinburgh, University of Toronto, University of California, Irvine, and University of Washington. Over decades, ICML intersected with milestones involving Geoffrey Hinton, Yoshua Bengio, Yann LeCun, Michael I. Jordan, Andrew Ng, Judea Pearl, Tom Mitchell, and Robert E. Schapire as machine learning matured alongside technologies from GPU manufacturers such as NVIDIA. The conference evolved amid parallel initiatives like ImageNet Large Scale Visual Recognition Challenge, GLUE benchmark, and projects at MIT Media Lab and Berkeley AI Research.

Paper submissions follow formats promoted by bodies like Proceedings of Machine Learning Research and use virtual tools from vendors such as OpenReview, CMT (Conference Management Toolkit), and platforms developed at Carnegie Mellon University. The review pipeline engages area chairs, senior program committee members, and external reviewers drawn from Google Brain, DeepMind, Facebook, Microsoft Research, Amazon Research, Adobe Research, Tencent AI Lab, and academia including Columbia University, University of Illinois Urbana-Champaign, Georgia Institute of Technology, University of Michigan, and University of Texas at Austin. Processes reference standards from Committee on Publication Ethics and incorporate reproducibility checks inspired by efforts at Journal of Machine Learning Research, Nature Machine Intelligence, Science Robotics, and Communications of the ACM.

ICML covers areas spanning supervised learning, unsupervised learning, reinforcement learning, probabilistic modeling, and deep learning, connecting to work by researchers at DeepMind on AlphaGo, AlphaZero, and reinforcement advances related to OpenAI Five. It includes subfields such as Bayesian methods with ties to Stanford Linear Accelerator Center researchers, kernel methods from Johns Hopkins University groups, causal inference influenced by Judea Pearl and institutions like University of California, Los Angeles, representation learning from Facebook AI Research, optimization techniques linked to Courant Institute of Mathematical Sciences, and applications in robotics from MIT CSAIL, NASA Jet Propulsion Laboratory, and Boston Dynamics. Topics interact with datasets and benchmarks like ImageNet, COCO, SQuAD, Penn Treebank, and tools such as TensorFlow, PyTorch, JAX, scikit-learn, and Keras.

ICML has hosted influential papers including advances in stochastic gradient methods, variational inference, generative models, and meta-learning. Landmark works presented by authors affiliated with Google DeepMind, OpenAI, Facebook AI Research, Stanford University, University of Toronto, and University of Montreal influenced follow-on projects like BERT, GPT, ResNet, Transformer (architecture), and Generative Adversarial Network. ICML publications have affected industry deployments at companies including Amazon, Alibaba Group, Baidu, Tencent, Salesforce, SAP, Siemens, and governmental research at Los Alamos National Laboratory and Lawrence Berkeley National Laboratory.

ICML rotates venues internationally, having met in cities such as Stockholm, New York City, Barcelona, Vienna, Beijing, Sydney, Montreal, San Diego, Tokyo, Paris, Athens, Vienna State Opera House adjacent locations, and conference centers tied to institutions like Royal Society and municipal centers. Attendance ranges from hundreds in early years to thousands recently, drawing participants from organizations including Google Research, Microsoft Research Cambridge, DeepMind London, Samsung Research, Huawei Noah's Ark Lab, LG AI Research, and academic labs at Imperial College London.

ICML presents awards and recognitions including best paper awards, outstanding paper recognitions, test-of-time awards honoring earlier contributions such as boosting and kernel methods, and student paper awards. Award recipients have included scholars associated with Turing Award winners and organizations such as Association for Computational Linguistics, IEEE Transactions on Pattern Analysis and Machine Intelligence, Royal Society, and recipients from National Academy of Engineering. ICML accolades have been cited alongside honors like NeurIPS Best Paper, AAAI Fellows, ACM Fellows, and national academies' memberships.

Category:Machine learning conferences