SUMMER SCHOOL ON MINING BIG AND COMPLEX DATA 04 - 08 September 2016 Ohrid, Macedonia

Program

The school will include lectures on predictive modelling methods for big and complex data. More specifically, the lectures will present methods handling the following complexity aspects: (a) structured data as input or output of the prediction process, (b) very large/massive datasets, with many examples and/or many input/output dimensions, where data may be streaming at high rates, (c) incompletely/partially labelled data, and (d) data placed in a spatio-temporal or network context. Each of these is a major challenge to current ML/DM approaches and is the central topic of active research in areas such as structured-output prediction, mining data streams, semi-supervised learning, and mining network data. The applicability and the potential of the presented methods will be demonstrated on several showcases from molecular biology, sensor networks, multimedia, and social networks.

The lectures will be given by world leading researchers and experts in machine learning and data mining, as well as experts from the application domains. The program will include talks on the following topics:

  • Semi-supervised learning for structured data
  • Kernel-based methods for structured data
  • Bayesian networks for multi-dimensional classification
  • Multi-label learning from batch and streaming data
  • Decomposition of the output spaces in structured output prediction
  • Architectures for distributed mining of big data
  • Mining data streams with structured outputs
  • Mining network data, network reconstruction and complex networks analysis
  • Analysis of streaming networks
  • Tensor data analysis
  • Spatio-temporal data mining
  • Redescription mining
  • Deep learning for image retrieval and image classification
  • Integrating deep learning with kernel methods
  • Controlling false discovery rates in multiple testing and selective inference
  • Gene function prediction and relating genotypes with phenotypes
  • Analysis of metagenomics data
  • Medical data mining
  • Analysis of neuro data and neuroimages

The program is crafted towards master and PhD-level students, and also professionals wishing to learn about cutting edge technologies for mining big and complex data.





Sunday 4 September 2016

09:00 - 09:45 | Saso Dzeroski: MAESTRA: Intro

09:45 - 10:30 | Dragi Kocev: Semi-supervised tree learning

10:30 - 11:00 | Coffee break

11:00 - 12:30 | Paolo Frasconi: Kernel methods for structured data

12:30 - 14:30 | Break

14:30 - 16:00 | Ulf Brefeld: Semi-supervised learning for SOP

16:00 - 16:30 | Coffee break

16:30 - 18:00 | Cesare Furlanello: Metagenomics data analysis

20:00 - Poster session



Monday 5 September 2016

09:00 - 10:30 | Jesse Read: Multi-label learning from batch and streaming data

10:30 - 11:00 | Coffee break

11:00 - 11:30 | Gjorgji Madjarov: Decomposition and structuring of the output space

11:30 - 12:00 | Pance Panov: Ontology of Data Mining

12:00 - 12:30 | Impromptu talk

12:30 - 12:45 | Break

12:45 - 19:30| Networking session at St. Naum Complex



Tuesday 6 September 2016

09:00 - 09:45 | Michelangelo Ceci: Mining network data

09:45 - 10:30 | Ljupco Todorovski: Network reconstruction

10:30 - 11:00 | Coffee break

11:00 - 11:45 | Hadi Fanee: Mining tensor data

11:45 - 12:30 | Ljupco Kocarev: Complex Networks Analysis

12:30 - 13:30 | Break

14:30 - 15:30 | Joao Gama: Mining streaming data and networks

15:30 - 16:00 | Saso Dzeroski: SOP on data streams

16:00 - 16:30 | Coffee break

16:30 - 18:00 | Albert Bifet: Architectures for distributed mining of big data

20:00 - Poster session



Wednesday 7 September 2016

09:00 - 09:30 | Michelangelo Ceci: Network Applications

09:30 - 10:15 | Donato Malerba: Spatio-temporal data mining: Part 1

10:15 - 10:45 | Annalisa Appice: Spatio-temporal data mining: Part 2

10:45 - 11:00 | Coffee break

11:00 - 12:30 | Ondrej Chum: Large Scale Image Retrieval and Mining

12:30 - 14:30 | Break

14:30 - 15:15 | Julien Mairal: Sparse estimation and dictionary learning

15:15 - 16:00 | Julien Mairal: Towards deep kernel machines

16:00 - 16:30 | Coffee break

16:30 - 18:00 | Ivica Dimitrovski: Deep learning for plant identification

18:30 - 20:30 | Walk in Ohrid

20:30 - School Dinner in Ohrid



Thursday 8 September 2016

09:00 - 10:30 | Fran Supek: Gene function prediction and relating genotypes with phenotypes

10:30 - 11:00 | Coffee break

11:00 - 12:30 | Myra Spiliopoulou: Mining medical data

12:30 - 14:30 | Break

14:30 - 16:00 | Yoav Benjamini: Controlling false discovery rates in multiple testing and selective inference and analysis of neuro data

16:00 - 16:30 | Coffee break

16:30 - 17:15 | Tomislav Smuc: Redescription mining and analysis of neuro data

17:15 - 18:00 | Saso Dzeroski: Closing remarks

18:00 - Farewell Party

Speakers

Registration

The MAESTRA Summer School is not a commercial event and payment of a registration fee is not required from the attendees. Instead, the major organization costs are covered by the FP7 EU project MAESTRA. The registration for the event covers admission to all sessions, school materials, coffee breaks, the school dinner, and the networking session at the St. Naum complex. The registration does not cover accommodation and meals (except the school dinner). The school venue offers lodging at very affordable rates (i.e., full-board service for 35 EUR per day, see the pricelist).

The venue provides limited seating, hence, the school will be open for 50 participants.

The registration for the summer school is now closed.

Organizers

Program Organizers

Sašo Džeroski, Dragi Kocev

Jozef Stefan Institute – JSI
Department of Knowledge Technologies

Michelangelo Ceci

University of Bari Aldo Moro – UNIBA
Departmet of Computer Science

Tomislav Šmuc

Rudjer Boskovic Institute – RBI
Computational Biology and Bioinformatics Group

Joao Gama

INESC Technology and Science – INESC TEC
Associated Laboratory of Institute for Systems and Computer Engineering

Ivica Dimitrovski

University "Ss. Cyril and Methodius" – UKIM
Faculty of Computer Science and Engineering

Local Organizers

Ivica Dimitrovski

Gjorgji Madjarov

Gjorgji Strezoski

Dario Stojanovski

University "Ss. Cyril and Methodius" – UKIM
Faculty of Computer Science and Engineering

Venue

Ohrid is within driving distance to Skopje, Thessaloniki and Tirana making it nicely connected to western Europe with budget flights through Wizzair in Skopje or EasyJet in Thessaloniki. The roads in general in Macedonia are safe and good. The main international highway is E-75 connects all major the cities to Skopje.
Most likely there will be options for organized transportation from Skopje airport to the venue site. Taxi fares are fairly cheap and start at approximately 20 - 50 MKD (0.33 - 0.80 EUR) and charge additional 25 MKD (0.4 EUR) per 1 km.

  • Skopje - Ohrid 170km

  • Thessaloniki - Ohrid (via Bitola) 280km

  • Tirana - Ohrid 130km

  • Sofija - Ohrid 320km