BME VIK - Algoritmikus tőzsdei folyamat-előrejelzés

A tantárgy neve magyarul / Name of the subject in Hungarian: Algoritmikus tőzsdei folyamat-előrejelzés

3. Course coordinator and department Dr. Telcs András,

4. Instructors

Name:	Position:	Department.:
László Györfi DSc	Professor	Department. of Computer. Science and Information Theory
András Telcs DSc	Assoc. Prof.	Department. of Computer. Science and Information Theory

5. Required knowledge

Recommended: Mathematical statistics, Finance, Planning of financial investments

7. Objectives, learning outcomes and obtained knowledge a Objectives, learning outcomes and obtained knowledge

The course provide knowledge of methodology of modeling and predicting financial time series and the related portfolio strategies.

b. Acquired skills

By the competition of the course students are enabled to apply tools, techniques to model and forecast financial time series. Will be able to support bank, investment founds in the planning their investment strategies.

8. Synopsis a Lectures

1. Estimate of the density function, L1 error.

2. Estimate of the density function, histogram.

3. Estimate of the density function, kernel estimates.

4. The regression problem, the regression function, partition method.

5. The regression problem, the regression function, kernel functions.

6. The regression problem, nearest neighbor estimate.

7. The regression problem, empirical error.

8. Pattern recognition, error probability.

9. Pattern recognition, Bayes decision, partitions.

10. Pattern recognition, kernel functions, nearest neighbor estimate.

11. Pattern recognition, empirical error.

12. Optimal portfolio strategies, fixed portfolios.

13. Optimal portfolio strategies, constantly rebalanced portfolios.

14. Optimal portfolio strategies, dynamically rebalanced and empirical portfolios.

b. Labs

1. Data acquisition, database design.

2. Data cleaning and cleansing.

3. Introduction to the software usage.

4. Exploratory data analysis, descriptive statistics.

5. Exploratory data analysis, presentation, graphics.

6. Time series analysis, overview, deterministic models.

7. Test.

8. Time series analysis, overview, application of ARIMA, ARCH, GARCH models.

9. The regression problem, the regression function, elementary, spline, NN.

10. The regression problem, the regression function, kernel functions – Gaussian.

11. The regression problem, nearest neighbor estimate implementation.

12. Test.

13. Log-optimal portfolio, algorithmic implementation static, rebalanced.

14. Log-optimal portfolio, algorithmic implementation combined experts.

9. Method of instruction Lectures and laboratory

10. Assessment a. Active involvement during lectures, condition of the course signature minimum, completion os lab exercises, 40% score of the two test average. Exam 70%, lab 30% in the final mark. The second test covers the first 11 lab and lecture material.

b. Oral exam

c. Pre exam subject of tutor's agreement

11. Recaps One re-test during the recap-weak and oral presentation 30% of missing lab exercises are possible.

12. Consultations upon personal inquiry

13. References, textbooks and resources 1.      Száz János: Tőzsdei opciók vételre és eladásra, Tanszék Kft, 1999.

2.      R. S. Tsay: Analysis of Financial Time Series, Wiley, 2nd edition, 2005.

3.      L. Györfi, M. Kohler, A. Krzyzak, H. Walk: A Distribution-Free Theory of Nonparametric Regression, Springer-Verlag, 2002.

4. L. Györfi, G. Ottucsák: Empirical log-optimal portfolio selections: a survey, http://www.szit.bme.hu/~oti/portfolio/articles/tgyorfi.pdf 2007

14. Required learning hours and assignment

classes	70
preparation for classes	10
preparation for labs	14
preparation for test	36
preparation for seminars	0
Vizsgafelkészülés	50
Összesen	180

15. Syllabus prepared by

Name:	Position:	Department:
László Györfi DSc	Professor	Department. of Computer. Science and Information Theory