Yandex, a NASDAQ-listed internet giant, has launched the world’s first free online data labelling course on the Coursera platform to address the increasing use of artificial intelligence in businesses.
Register for our upcoming Virtual CxO Round Table | 18th Mar | 5 pm – 6:30 pm
The course titled — Practical Crowdsourcing for Efficient Machine Learning by Yandex is available free of charge on the Coursera platform from March 15. Taking a hands-on approach, the course will be taught by professionals from Toloka, a successful crowdsourcing data labelling platform, stated in the official release. Toloka will present real-life case studies on labelling and processing data, enabling students to master crowdsourcing concepts and even launch their own data labelling projects during the five-week course.
While AI’s use becomes more commonplace, the advancement of the field still relies on human intelligence because preparing data for AI and machine learning requires people to label the data. Much of this preparation consists of data labelling and usually involves text or image classification or checking audio transcriptions. Thus, according to a report, the market for third-party data labelling is forecast to surpass $1 billion by 2023 from just $150 million in 2018.
Commenting on this new launch, Olga Megorskaya, the CEO of Toloka, commented — as the pandemic accelerates digital adoption and fuels the demand for tech talent, mastering crowdsourcing technologies can give individuals and businesses the required competitive edge. This new course will enable students to apply the theory they learn to real-life machine learning tasks and projects.
Crowdsourcing has already been a core part of tech companies and R&D powerhouses worldwide, including NASA. Yandex also teaches courses on crowdsourcing at the Higher School of Economics, Moscow Institute of Physics and Technology, and in Russia’s leading research universities. As a matter of fact, crowdsourced solutions fill a gap in the market, where the demand for scalable, cost-efficient and high-quality data labelling is growing. It is based on breaking down a large, complex job into small, clearly-defined tasks that can be handled by non-experts, who can swiftly gain the skills they need to complete these tasks well.
Some elements of this newly introduced course on Coursera have already been presented and validated in the form of workshops and tutorials at several international machine learning and data management events, such as NeurIPS 2020 and CVPR 2020. However, the entire course is now available on Coursera to anyone worldwide, with access to the internet.
No background knowledge is required to avail of this course. Still, it would help professionals like ML developers, data analysts, researchers, and students, who are just considering these careers.
Get more information about this course here.