Atlas ML

AtlasML is a London based technology company building an open-source resource for machine learning platform and tools. The company aims to make machine learning and artificial intelligence technology accessible to anyone for building deep learning projects and applications.

Atlas ML is an artificial intelligence company making open source tools for deep learning that is headquartered in London, United Kingdom. Their main community product is Papers With Code, which is the world's largest machine learning resource for models and results.

AtlasML operates a free web-based resource for deep learning papers and code called “Papers with Code.” The website holds resource for machineCode", learningcontaining research codepapers, implementations, researchof papers, and evaluation tables. The contents of the site are licensed under Creative Commons (CC BY-SA), which is the same as Wikipedia.

Content on the site is partly automated and partly crowd-sourced, and; researchers can submit code implementation and results for a specifictheir categorypapers. As of April 2019, the platform has parsed over 60,000 research papers to categorize them as method and tasks. ArXiv abstract is used to annotate tasks and datasets manually. The data is sourced through manual annotation of research papers or extracted with the company’s open source scraper, SOTA extractorcategorised pipeline.in Registeredvarious usersmachine canlearning editfields and add code, tasks, and evaluation results for available papers on the website. The “Papers with Code” community uses Slack for real-time discussions.

Robert Stojnic and Ross Taylor founded the company on June 1, 2018. Robert was one of the core developer of Wikipedia from 2006-2010 with experience in Machine Learning, Natural Language Processing, and Data Science. Ross is a machine learning specialist with expertise in predictive and quantitative analysis of financial markets.

Robert Stojnic and Ross Taylor founded the company on June 1, 2018. Robert was a core developer at Wikipedia from 2006-2010 with experience in Machine Learning, Natural Language Processing, and Data Science. Ross was a quantitative analyst designing machine learning trading strategies in several firms, and the author of one of the most popular Python time series libraries, with experience in Machine Learning, Time Series Methods and Approximate Inference.