Industry Webinar: Human and ML — Collaboration in Data Labeling

In this webinar, we will consider an approach to the formation of high-quality data using a scalable crowdsourcing approach. Data quality in the current pipeline of creating AI products is an important component along with computing power (such as GPU, TPU, etc) and “algorithms” for data transformation (in fact, machine learning methods). However, stable markup of high-quality data is a complex and not scalable approach. We will look at how you can combine knowledge from current state-of-the-art models to aggregate, construct the correct answer, create dynamically calculated prior knowledge about the correct labels, apply human verdicts to improve and accelerate the receipt of correctly annotated data.

Join Evgenii Sorokin (Machine Learning Engineer from Toloka) on Wednesday, December 15, to learn more about human and ML collaboration on data labelling. Participants will have an opportunity to ask questions, so come prepared to engage!

Please register following the link