Магистратура
2022/2023
Машинное обучение I
Кто читает:
Департамент экономики
Когда читается:
1-й курс, 3, 4 модуль
Формат изучения:
с онлайн-курсом
Онлайн-часы:
72
Охват аудитории:
для своего кампуса
Язык:
английский
Кредиты:
6
Контактные часы:
32
Course Syllabus
Abstract
Using data to make predictions, test hypotheses and estimate models is an important skill in the current job market. Many companies collect a lot of data and their decisions data-driven. Machine learning disrupts many fields and promises to achieve superhuman performance in the coming decades. Statistical analysis allows to test hypotheses and verify which of the models fits the data best. In this course we will cover different methods for supervised and unsupervised learning to develop a necessary toolkit for successful data scientists. For some of the methods we will go into details to learn why and how they work. Also we will touch on ethical implications of data science in the age of big data and apply learned methods to real business data sets.
Learning Objectives
- Students will free comfortable orienting among different methods of machine learning and develop understanding of why these methods work and to extend them
Expected Learning Outcomes
- Understand the concept of data generating process and how it is different to the concept of model
- Learn more details on hypothesis testing
- Understand different methods for supervised learning such as regressions, random forest, gradient boosting etc.
- Able to design and implement efficient feature engineering processes, including working with gaps, categorical variables, time series, and text
- Can create reproducible scripts and pipelines for data processing and model training in Python
- Know how to prepare data for training models, including scaling, normalization, and sampling
Course Contents
- Python for data science
- Supervised machine learning
- Unsupervised machine learning
- Machine learning principles: cross-validation, feature selection, metrics
- Tools for data science
Bibliography
Recommended Core Bibliography
- Silver, N. (2012). The Signal and the Noise : Why So Many Predictions Fail-but Some Don’t. New York: Penguin Books. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=1122593
Recommended Additional Bibliography
- Bruce E. Hansen. (2013). Econometrics. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsbas&AN=edsbas.C0DB9E1E