Yandex NLP week

26 – 29 марта
Organised by the Yandex School of Data Analysis and Yandex.Research


NLP Week at Yandex is a research-oriented intensive course on Natural Language Processing. This four-day course is an extension to the NLP course taught at the Yandex School of Data Analysis. It explores some of the advanced ideas that go beyond the YSDA program and is offered to students with some experience in Natural Language Processing.

The course lecturers are Mirella Lapata, Professor in Natural Language Processing at the School of Informatics of the University of Edinburgh, and Wilker Aziz, Assistant Professor in Computational Linguistics at the Institute for Logic, Language and Computation at the University of Amsterdam.

This course will be taught in English.

What is this course about?

  1. Latent Variables in NLP. Basic concepts, relation to classic probabilistic models in NLP. Deep generative models with continuous variables. Deep generative models with discrete variables. Applications to NLP.

  2. Semantic parsing. Mapping a natural language sentence into a formal representation of its meaning is rife with problems for Natural Language Processing. We will talk about approaches to solving the problem of ambiguity and extracting implicit meanings.

Who is this course for?

NLP Week at Yandex is designed for:

● Students of the Yandex School of Data Analysis who have completed the NLP course

● Undergraduate and graduate students with a strong background in mathematics and a decent knowledge of Machine Learning and Deep Learning interested in Natural Language Processing

● Researchers and specialists working in NLP or related areas wishing to expand their knowledge and skills

What background knowledge do you need?

● Natural Language Processing: knowledge of and experience in NLP; completion of the NLP course at the YSDA or some other NLP course.

● Machine Learning: strong knowledge of Machine Learning and Deep Learning.

● Mathematics: proficiency in linear algebra and probability theory is highly desirable.

● Programming: Python and familiarity with PyTorch.

● English: strong understanding of professional English.


  • 26.03, 18:00 - 21:00 – Latent variable models and posterior inference (Wilker Aziz)
  • 27.03, 18:00 - 21:00 – Deep generative models for NLP (Wilker Aziz)
  • 28.03, 18:00 - 21:00 – Advanced topics (Wilker Aziz)
  • 29.03, 18:00 - 21:00 – Semantic parsing (Mirella Lapata)