MPF_AIIF AI in Finance

Faculty of Economics and Administration
Autumn 2023
Extent and Intensity
2/2/0. 6 credit(s). Type of Completion: zk (examination).
Teacher(s)
prof. Ing. Štefan Lyócsa, PhD. (lecturer)
Ing. Martina Halousková (assistant)
Guaranteed by
prof. Ing. Štefan Lyócsa, PhD.
Department of Finance – Faculty of Economics and Administration
Contact Person: Iva Havlíčková
Supplier department: Department of Finance – Faculty of Economics and Administration
Timetable
Fri 8:00–9:50 P106, except Fri 22. 9., except Fri 10. 11.
  • Timetable of Seminar Groups:
MPF_AIIF/01: Fri 12:00–13:50 VT204, except Fri 22. 9., except Fri 10. 11., Š. Lyócsa
MPF_AIIF/02: Fri 10:00–11:50 VT206, except Fri 22. 9., except Fri 10. 11., Š. Lyócsa
Prerequisites
Students are expected to be familiar with basic concepts of Statistics and Econometrics.
Course Enrolment Limitations
The course is only offered to the students of the study fields the course is directly associated with.

The capacity limit for the course is 35 student(s).
Current registration and enrolment status: enrolled: 10/35, only registered: 1/35
fields of study / plans the course is directly associated with
Course objectives
The ever increasing amount of data already dictates our understanding of finance. The ability to process, study, interpret and present such data leads to an enormous competitive advantage on the job market; the Machine Learning in Finance gives students the possibility to work towards this advantage. The course is centered on key topics of machine learning with specific emphasis on case studies applied in the context of financial markets, credit and profit scoring, hedonic price models for real estate and used cars. Key topics include: data pre-processing, unsupervised learning methods, predictive modelling via OLS, LASSO, RIDGE, EN, Complete Subset Regressions, Logistic regression, Random Forest. Basic principles of Gradient Boosting, Support Vector Machines or other methods are also discussed. Selected principles discussed in the course are handling of data-snooping bias, hyper-parameter tuning, bagging and boosting, ensemble learning. Course is primarily led in program R.
Learning outcomes
After completing the course, the student should be able to: - identify areas where data techniques might be useful, - prepare and design data analysis, - understand key concepts of machine learning, - program and present analysis using R or Python.
Syllabus
  • Introduction 1. Introduction to Artificial Intelligence in Finance – applications in Finance, variance bias trade-off, supervised and unsupervised learning, semi-supervised learning, reinforced learning, critique of machine learning with over-fitting and model interpretability. 2. Data pre-processing – handling missing data part I., outliers, data transformation, feature engineering. Supervised AI in Finance - continuous outcome 3. Standard machine learning framework – multivariate regression, interactions, dummies, model/variable selection, in-sample and out-of-sample approach, loss functions for continuous target variables. Data snooping bias – model confidence set. 4. Regularization techniques – LASSO, Ridge, Elastic net. hyper-parameter tuning – cross-validation, leave-one-out, grid search, Complete subset regressions, Decision trees, Random forest and support vector machines. 5. Tree-based methods - decision trees, pre-pruning, post-pruning, bagging, random forest, boosting trees. Supervised AI in Finance - discrete outcomes 6. Logistic regression – marginal effects. Discrete choice model evaluation – confusion matrix. AUC. 7. LASSO, Ridge, Elastic net, Complete Subset Logistic Regression. 8. Tree-Based methods, decision trees, pre-pruning, post-pruning, bagging, random forest, boosting for discrete outcomes. Unsupervised machine learning 9. Unsupervised machine learning – distance measures, k-means, k-medoids, CLARA, agglomeration clustering, cluster validation and optimization. 10. Principal component analysis, network based feature selection techniques. Further topics in machine learning 11. Forecast combination - ensemble techniques, Time-series models, model based approach to missing data, sample selection. 12. Further modelling options: Linear Discriminant Analysis, Naïve Bayes classification. 12. Further modelling options: Support vector machines, Neural Networks.
Literature
    recommended literature
  • COQUERET, G and Guida T AMP. Machine Learning for Factor Investing: R Version. CRC Press, 2020, 341 pp. ISBN 978-0-367-54586-4. info
  • CLASTER, W B. Mathematics and Programming for Machine Learning with R: From the Ground Up. CRC Press, 2020, 430 pp. ISBN 978-0-367-56194-9. info
  • NWANGANGA, F and M CHAPPLE. Practical Machine Learning in R. Wiley, 2020. ISBN 1-119-59151-1. info
  • WILEY, Matt and Joshua F. WILEY. Advanced R statistical programming and data models : analysis, machine learning and visualisation. California: Apress, 2019, xx, 638. ISBN 9781484228715. info
  • LANTZ, Brett. Machine learning with R : expert techniques for predictive modeling. Third edition. Birmingham: Packt, 2019, xiii, 437. ISBN 9781788295864. info
  • BERNARDI, M and Catania L AMP. The model confidence set package for R. International Journal of Computational Economics and Econometrics. 2018, vol. 8, No 2, p. 144-158. Available from: https://dx.doi.org/10.1504/IJCEE.2018.091037. info
  • WRIGHT, M N and A ZIEGLER. ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R. JOURNAL OF STATISTICAL SOFTWARE. LOS ANGELES: JOURNAL STATISTICAL SOFTWARE, 2017, vol. 77, No 1, p. 1-17. ISSN 1548-7660. info
  • ELLIOTT, G, A GARGANO and A TIMMERMANN. Complete subset regressions. Journal of Econometrics. 2013, vol. 177, No 2, p. 357-373. Available from: https://dx.doi.org/10.1016/j.jeconom.2013.04.017. info
    not specified
  • JAMES, Gareth R., Daniela WITTEN, Trevor HASTIE and Robert TIBSHIRANI. An introduction to statistical learning : with applications in R. Second edition. New York: Springer, 2021, xv, 607. ISBN 9781071614174. info
Teaching methods
Lecture notes, problem sets, and case studies are necessary for successful passing of the course. As they are required they will be available in the eLearning module.
Assessment methods
Grading is in accordance with the internal guidelines of the Faculty of Economics and Administration of Masaryk’s University and is based on midterm (30%) and final exam (70%). A total of 51% minimum is required to pass. MUNI students who study abroad have to pass the midterm and the exam.
Language of instruction
English
Further Comments
Study Materials
The course is taught annually.
The course is also listed under the following terms Autumn 2022, Autumn 2024.
  • Enrolment Statistics (Autumn 2023, recent)
  • Permalink: https://is.muni.cz/course/econ/autumn2023/MPF_AIIF