• A
  • A
  • A
  • АБB
  • АБB
  • АБB
  • А
  • А
  • А
  • А
  • А
Обычная версия сайта
Бакалавриат 2021/2022

Экономические приложения машинного обучения

Статус: Курс по выбору (Экономика)
Направление: 38.03.01. Экономика
Когда читается: 4-й курс, 1 модуль
Формат изучения: без онлайн-курса
Охват аудитории: для своего кампуса
Язык: английский
Кредиты: 3

Course Syllabus

Abstract

The course describes main recent machine learning and data analysis methods as well as their application in economic research. Special attention in the course is paid to the implementation of these algorithms and models in Python. All lectures will be held online in zoom, and seminars will be mixed online-offline. There will we 3 individual written assessments and 1 group oral assessment. Problem sets submitted before the deadline weight 1:1. Problem sets submitted after the deadline weight 2:1. If a student does not get a passing grade by the end of the course, there will be 2 makeups.
Learning Objectives

Learning Objectives

  • Within this course students learn the key methods of data analysis, examples of their application to economic research and learn how to build on their own considered models in Python.
Expected Learning Outcomes

Expected Learning Outcomes

  • Know where to find data, understand the formats, know how to work with it and do simple operations on it.
  • Know how to run and visualize a regression. Write an OLS regression from scratch.
  • Code a Logit regression from scratch, run a classic Logit regression in Python, know alternative Logit regressions.
  • To understand decision trees, forests; run them in Python and present the results.
  • Know and distinguish various non-parametric methods, such as KDE and RD.
  • Understand how Logit fits into a broader family of classification methods.
  • Know various SVM methods.
  • Know how to run a gradient boosting.
  • Know how to run a simplest neural network.
Course Contents

Course Contents

  • 1. Data
    1.1. Alternative to the official data sources and new ways of data collection for the analysis. Organizational issues and information on the group projects. 1.2. API vs dumps vs scraping in Python. Data storage formats: tables vs json
  • 2. Regressional and visual analysis
    2.1. Regressions with regularization (Ridge, lasso, elastic net regression). Cross-validation. Examples of applications in economic research. 2.2. Data types and computer arithmetics limits. Data visualization
  • 3. Logit
    3.1. Binary choice. Types of logit models and application areas: nested, ordered, random logit. 3.2. Convex vs non-convex optimization. Calculation acceleration: cycles vs matrices. Monte-Carlo simulation
  • 5. Classification
    5.1, 5.2. Classification. Examples of its implementation. Practice.
  • 6. Support vector machine
    6.1, 6.2. SVM. Main aspects of building SVM in Python.
  • 4. Nonparametric analysis
    4.1. Probability estimation (Nadaraya-Watson), discontinuities (Regression Discontinuity), and monopoly pricing (Guerre-Perrigne-Vuong) 4.2. Nonparametric methods, data heterogeneity
  • 7. Decision trees
    7.1, 7.2. Decision trees, random forest. Examples of their use in economic research. Practical part.
  • 8. Decision Trees
    8.1, 8.2. Gradient boosting. Boating based on decision trees. Practical part.
  • 9. Neural networks
    9.1, 9.2. Artificial Neural networks. Key principles and model characteristics. Examples of the use of neural networks in economic research.
Assessment Elements

Assessment Elements

  • non-blocking Контрольная 1 (Письменно)
  • non-blocking Контрольная 2 (Письменно)
  • non-blocking Контрольная 3 (Письменно)
  • non-blocking Групповая презентация (в конце модуля)
Interim Assessment

Interim Assessment

  • Interim assessment (1 module)
    0.25 * Групповая презентация (в конце модуля) + 0.25 * Контрольная 1 (Письменно) + 0.25 * Контрольная 2 (Письменно) + 0.25 * Контрольная 3 (Письменно)
Bibliography

Bibliography

Recommended Core Bibliography

  • Дмитриев Егор Андреевич. (2017). Линейная регрессия. Students’ Scientific Research and Development ; № 2(4) ; 123-124 ; Научные Исследования и Разработки Студентов.
  • Красногир, Е. Г. (2009). Непараметрические ядерные оценки Надарая–Ватсона и область их задания.

Recommended Additional Bibliography

  • Aguirregabiria, V., & Carro, J. M. (2021). Identification of Average Marginal Effects in Fixed Effects Dynamic Discrete Choice Models.
  • Wiktor Budziński, & Mikołaj Czajkowski. (2021). Accounting for Spatial Heterogeneity of Preferences in Discrete Choice Models. Central European Journal of Economic Modelling and Econometrics (CEJEME), 13(1), 1–24. https://doi.org/10.24425/cejeme.2021.136456