Data Science Winter School

Data Science Winter School 2018 is organized by the Faculty of Applied Science of the Ukrainian Catholic University and is a part of the sequence of the winter and summer schools. This winter the school’s participants – master and Ph. D. students, young professionals – could take one of the two courses:  Reinforcement Learning and Time Series & Forecasting. The participants will gain solid theoretical knowledge as well as a lot of practical work.

Please notice that participants’ knowledge and skills level should be intermediate or higher. The students have to be familiar with the linear algebra, the theory of probability, and statistics.

The school’s dates are January 10 – 17, 2018.  Please notice, the proposed courses will go simultaneously. Thus the one has to pick the only one of them.

School schedule

  • January 10  – check-in
  • January 11-13 – studying
  • January 14 – day off
  • January 15-17 -studying

 

A study day on each course has the following structure of the schedule:

  • theoretical lecture part before lunch
  • practical work under lecturers’ supervision after lunch

 

The application dates – December 1 for early birds and December 15 for the late night owls.
The location is the Ukrainian Catholic University, Lviv, Ukraine.

Study language – English. The participants will gain 3 ECTS.

 

Application is closed

 

Reinforcement Learning

Course key topics:
– Bandit algorithms
– Markov decision processes
– Model-free control
– Value function approximation
– Policy gradients
– Meta-learning
– Learning through self-play
– Deep reinforcement learning

More information on the Reinforcement Learning course web page.

Lecturer

pablo_maldonado

Pablo Maldonado, Ph.D.

Applied mathematician and data scientist consultant
Czech Technical University, Prague

Time Series & Forecasting

Course key topics:
– Typical patterns in time series data
– Time series decomposition
– Forecasting
– Forecasting using regression
– ARMA modelling
– ARCH/GARCH: models with conditional volatility
– Basic models for multivariate time series data
– Forecasting special data

More information on the Time Series & Forecasting course web page.

Lecturer

pablo_maldonado

Dr. Yarema Okhrin

Professor of Statistics
at the University of Augsburg, Germany

Participation fee

The participation fees are presented in the table below. Please notice the application deadlines. The participation fee doesn’t include accommodation or meals.

option december 1  DECEMBER 15
Ukrainian participants 8 000 UAH 10 000 UAH
International participants 300 Euro 350 Euro

 

Discounts

Organizers can provide discounts which cover up to 30% of tuition fee for bachelor, master, and Ph.D. full-time students from the Ukrainian universities. In order to apply for a discount, the eligible students should fill an additional form and:

  • provide the personal statement that explains a discount need;
  • provide a letter of reference from a teacher of the home university.

 

A number of discounts are limited. The organizers are not obligated to explain their decision in a case of the acceptance/rejection of a discount application. Apply for a discount by the link: https://goo.gl/7rTTuV.

 

Prerequisites

Eligible applicants should have solid knowledge of linear algebra, the theory of probability, and statistics; should have good Python knowledge and R (optimal for the Time Series & Forecasting course). The specific prerequisites are mentioned on a separate page of each course.

 

Application process

Before applying for the workshop(s) one should become familiar with the terms of service, application rules, fee payment procedure, study schedule, certificate issuance, accommodation resources. Please, follow the “Terms of service” web page to read more.

Attention! Due to the high amount of the application, we had to close the enrollment sooner than the late deadline.

 

Application is closed

 

Previous schools and workshops

 

Contacts

In case you have any questions, please contact us via