The goal of the course is to integrate the knowledge students have acquired while studying theoretical and computer disciplines into a unified system, to develop practical skills in working with modern computational linguistics tools, and to build experience in research work.
Expected learning outcomes:
1. Students will learn to identify current problems in the field of computational linguistics and formulate project tasks.
2. They will complete the full project lifecycle: from initial data preprocessing to result analysis and interpretation.
3. They will gain experience working with large datasets and analyzing textual information using modern tools.
4. They will develop skills in critically evaluating both their own projects and the results of other researchers, conducting testing, and assessing the quality and accuracy of developed systems.
5. They will enhance their skills in public speaking and scientific communication during the presentation and defense of their own projects.
The course is divided into three modules:
- The first module focuses on describing the tasks students will be working on.
- The goal of the second module is to create a baseline model for the target task.
- The third module aims to improve the baseline model.
To successfully complete the course, students need knowledge of:
- automatic text processing
- linear algebra and mathematical analysis
- probability theory and mathematical statistics
- classical machine learning methods
- the basics of deep learning
Students are also expected to be able to program in Python.
Topics covered in the course
Topic 1. Introduction to machine learning.
Topic 2. Reinforcement learning and deep learning.
Topic 3. Training dataset: principles of data collection and synthetic data generation.
Topic 4. Baseline model and error analysis.
Topic 5. Model selection: pipelines.
Topic 6. Feature generation and fine‑tuning.
Topic 7. Metrics and criteria for model evaluation.
Topic 8. Model optimization methods.
Classes are held at the Faculty of Philology, Lomonosov Moscow State University, 1st Humanities Building, Room 950
The course program includes 18 sessions: 10 lectures and 8 seminars
Format: offline
Course start date: February 10, 2025
Class schedule:
- Third and fourth periods (from 13:00 to 16:10) on February 10 and 24; March 10, 17, 24, and 31; and April 7.
- Fifth and sixth periods (from 16:20 to 18:30) on February 18 and March 4.
Course Telegram channel
Enrollment for the 2025 course is closed