Statistical Learning Using R

Statistical Learning Using R

The main aim of the course is to deliver the knowledge of the basics of statistical learning. The first part of the course will deal with classical techniques as comparison of batches, decision trees, random forests, bagging, boosting, k-means and clustering classification.

On the second day we will enter the world of the multivariate distributions based on copulas. Making a way over bivariate models we deeply discuss the Hierarchical Archimedean Copulas. Together we will apply the methods to the true and simulated data using R statistical software. Prerequisite is the knowledge of the elements of matrix algebra and understanding of random variables and distributions functions.

Event Details

3 - 4 October 2018
9:00 am - 12:00 pm
FBE PC Lab 206 (Wednesday) and PC Lab 306 (Thursday)
Building 6ER
Friday 28 September 2018

The Instructor

Professor Ostap Okhrin is the head of the Chair of Econometrics and Statistics, specifically in the Transport Sector at Technische Universität Dresden (TUD). Professor Okhrin’s research concentrates on the interface between statistics, probability theory and finance with strong econometric elements. His current research activities can be split into several main areas around mainly three FOR codes, Statistics (0104), Finance (1502) and Econometrics (1403).

The main research activity, which successfully led to projects like B10 project within the scope of SFB649 (HU Berlin), great!ipid4all DAAD (TU Dresden)[1] , covers the methods of modelling dependency structures and their applications in Finance, Economics, Environmental Finance and other fields.

Okhrin, Okhrin and Schmid (2013) is among 5 most cited papers in the Journal of Econometrics (ABDC A*) since 2013. The development of new copula families, which are appealing for practical applications, investigation of their properties, and their application to different fields like portfolio theory, crop insurance, time series modelling, engineering, are one of the aspects of his research. Prof. Okhrin was the first who made steps towards copula-based models for high-frequency and spatial data.

He has published in top (ABDC A*/A) Journals in Statistics and Econometrics, like Journal of the American Statistical Association (ABCD A*), Journal of Empirical Finance (ABCD A), Journal of Econometrics (ABCD A*), Insurance: Mathematics and Economics (ABCD A), Journal of Risk and Insurance (ABDC A), Econometric Theory (ABCD A), Energy Economics (ABDC A*) etc.

All in all, he has published over 25 papers in the international peer-reviewed journals. Okhrin is also either leading or part of promising industry engineering research projects involving statistical modelling such as: Modelling of the top-of-descent for the air transportation (approved DFG proposal with Air Logistics, TU Dresden); Investigating the stability of the free parking location systems via wireless transmitters; Estimating the needed additional width of the Rhine River for the smooth traffic flow of the existing fleet (project funded by German Federal Institute of Water Operations).

Professor Okhrin is an Associate Editor of four highly ranked International Journals such as:Advances of Statistical Analysis, Computational Statistics, Dependence Modelling and Frontiers in Applied Mathematics and Statistics and served as guest Editor for two others. He has also published, in 2017, a book in Springer on “Basic elements in computational statistics”  with Y. Okhrin and W. Härdle. Prof. Okhrin past research visits include, ORFE Princeton University (2011), University of Chicago (2012), Southwestern University of Finance and Economics China (2016).


Back to the top of this page