Astrostatistics School Crete

About The Summer School

The goal of the Summer School for Astrostatistics is to guide advanced undergraduates, masters and graduate students, and early postdocs in the use of the most common statistical tools for the analysis of a wide variety of data. Each session includes an introductory class followed by a workshop (hands-on analysis of astronomical data), and poses particular emphasis on the development of coding abilities which will serve as a quick reference for everyday work.

The school is organized by the astrostatistics working group centered on the Department of Physics of the University of Crete.

The Department will provide the computational facilities required by the classes: the attendees do not need to use their personal computers. No previous experience with Python is required.

Where

Department of Physics, University of Crete, Heraklion

When

Tuesday to Friday
18-21 June 2019

Registration

Registration closed on 26 April 2019

Topics

Intro to Python and Jupyter Notebook
Linear regression
Classical statistical distribution tests, hypothesis testing
Bayesian statistics
Markov-Chain Monte Carlo
Machine learning classification
Machine learning regression
Machine learning clustering
Time series analysis

School Schedule

Please consult the School Schedule

Instructors

Teaching Body

Dr. Jeff Andrews

University of Copenhagen

Dr. Paolo Bonfini

Univ. of Crete, FORTH, National Obs. of Athens

Kostantinos Kovlakas

University of Crete

Invited Speakers & Teaching Assistants

Dr. Grigorios Tsagkatakis

Institute of Computer Science - FORTH

Prof. Davide Martizzi

Niels Bohr Institute

Dr. Grigoris Maravelias

National Observatory of Athens

School Venue

The school is hosted by the Department of Physics of the University of Crete, which will provide the computational facilities required by the classes.

Department of Physics, University of Crete, Heraklion

website

Directions

The Department is about 8 km from the center of Heraklion. There is a regular bus service from/to the city port (bus #20) or or airport (bus #11) to the University (direction Panepistimio / PAGNH). The municipality of Heraklion also provides an oline tool to keep track of the timetable and location of the buses (Iraklio Urban Buses), which is also available as mobile app. A detailed map and driving directions are available from the Department of Physics visitor information.

Heraklion airport is linked to the major European cities via Aegean Airlines, Olympic Air, easyJet, Ryanair, plus seasonal charter flights. Internal (Greece) flights are also provided by Ellinair and Sky Express. Daily ferry services from the Piraeus port in Athens are provided by Minoan Lines and Anek Lines. These ferries reach the port of Heraklion in the early morning, and leave from Heraklio in the late evening. This schedule gives the participants more than aboundant time to reach the School location in time for the first class, and to reach the port after the conclusion of the School.

Accomodation

June is considered touristic high season for Crete: attendees are advised to book their accommodation well in advance. To facilitate the search, in our subscription FORM we invite the participants to indicate whether they are prone to share a room.

Ideal locations to seek for an accomodation would be: 1) the old city center within the enceinte of the Venetian walls, or 2) along the line of the city bus route #11 or #20.

Rooms/Apartments

A variety of accomodations are rentable in the form of rooms or private apartments. As of the beginning of April, several options below 50€/night are still available through the common search engines, e.g. Booking or AirBnB, but the ones located in the city center will probably be fully booked soon.

Hotels

The hotels listed below provide an affordable accomodation in the city center. An additional list of hotels can be found on the visitor's page of the Department of Physics of the University of Crete.

Data

Workshop Jupyter notebooks

All the Jupyter notebooks presented at the school are publicly available at our github repository
Follow the instructions in the README file to download the relevant data sets
Deep Learning material

The slides from the Deep Learning seminar can be accessed here

The Deep Learning notebook regarding Galaxy Morphology classification is implemented through the Kaggle kernels, and can be accessed here
Using the Kaggle kernels will require you to sign in with one of your social media accounts, or to create one Kaggle account — once this is setup, just load the Galaxy Morphology kernel by following these simple instructions