Discover cutting-edge learning with Yandex Research
We are excited to present our array of educational courses designed by our researchers. We recently introduced a graph-based machine learning course at the Yandex School of Data Analysis. Additionally, we offer diverse programs in collaboration with partner universities and other research groups.
Graphs in Machine Learning
Yandex School of Data Analysis
This course explores machine-learning challenges involving graphs, introducing participants to effective data management strategies and a mix of classic and modern solutions. Students will learn to tackle diverse applications, such as molecular property prediction, combinatorial optimization, community detection, using advanced graph-based methods.
Deep Learning 2
HSE University
The DL2 course delves into a range of topics, including computer vision, transformers, generative models, graph neural networks, and tabular deep learning. Upon completing this course, many students successfully secure positions in the Yandex Research ML Residency program.
Machine Learning and Applications Research Seminar
HSE University
In this seminar, students analyze recent papers, embark on their initial research endeavors, and familiarize themselves with the latest advancements in the field.
Efficient Deep Learning Systems
HSE University
This popular course offers comprehensive coverage of experiment management and pipeline versioning fundamentals, training optimizations, distributed training, model parallelism, gradient checkpointing, offloading, sharding, and more.
Practical Deep Learning
HSE University
This course starts with deep learning fundamentals, from backpropagation and optimization methods to advanced techniques and PyTorch basics. By the third week, students learn convolutional neural networks and practical applications in computer vision.
Practical Reinforcement Learning
HSE University and Yandex School of Data Analysis
This course navigates the essentials of reinforcement learning, from foundational concepts like MDP and Bellman equations, to practical value- and policy-based algorithm families and deep reinforcement learning. The course centers around model-free RL algorithms prevalent in modern large-scale problems such as RLHF and robotics, but also covers model-based RL (planning), Imitation Learning, Inverse Reinforcement Learning, and more. Students explore practical applications and advanced techniques, gaining hands-on experience in reinforcement learning scenarios and methods.
Natural Language Processing
Yandex School of Data Analysis
The NLP course covers key areas in natural language processing, starting with word embeddings and moving on to text classification, language modeling, and sequence-to-sequence models with attention mechanisms. It includes transfer learning, focusing on powerful real-world models such as BERT and GPT. The latter half of the course explores recent advances in large language models and their applications, including prompting techniques, parameter-efficient finetuning, chain-of-thought reasoning, and more. The practical assignments range from training small NLP models from scratch to fine-tuning powerful open-source and open-access LLMs.
Course materials (github)
Deep Vision and Graphics
Yandex School of Data Analysis
This course offers an in-depth exploration of deep learning in visual computing. It covers neural network fundamentals, modern architectures based on convolutional networks and vision transformers, fine-tuning and transfer learning, and interpreting model predictions. The course covers practical applications including object detection, image segmentation, as well as generative modeling. In terms of generative modeling, this course explores a range of techniques from GANs and variational autoencoders to modern diffusion-based methods.