12th International Symposium on Intelligent Distributed Computing
IDC 2018
15-17 October 2018, Bilbao, Spain

Sponsored by:

Latest news:

October 14th, 2018:
Slides for the opening talk available.

October 13th, 2018:
Social Program available (further details to be given during the conference opening)

October 11th, 2018:
New invited speaker: Josu Ceberio, in replacement of Prof. Herrera (apologies on his behalf due to illness)

September 12th, 2018:
Early registration fees will be maintained until the celebration of the conference

September 4th, 2018:
Program is available

June 28th, 2018:
Registration is open

June 4th, 2018:
Updated notification deadline: June 5th, 2018

May 1st, 2018:
Submission deadline extended: May 15th, 2018 (no more extensions will be granted)

April 9, 2018:
Submission deadline extended: May 1st, 2018.

March 26, 2018:
New tutorial: ANDROPYTOOL.

March 6, 2018:
Confirmed Special Issue on Applied Soft Computing.

March 6, 2018:
New invited speaker: Albert Bifet.

January 19, 2018:
New Accepted Workshop: ML-PdM.

December 13, 2017:
New tutorial: JMETALSP.

December 13, 2017:
New tutorial: KMBD.

December 7, 2017:
New invited speakers: Francisco Herrera and Eleni I. Vlahogianni.

December 5, 2017:
New Accepted Workshop: COMPSUS.

November 15, 2017:
New Accepted Workshop: INDILOG.

November 6, 2017:
Confirmed Special Issue on Future Generation Computer Systems.

October 30, 2017:
Definitive conference dates published.

October 25, 2017:
Tentative conference dates published.

October 20, 2017:
First CFP published.

October 19, 2017:
Invited speakers: Jose A. Lozano and David Camacho.

October 18, 2017:
IDC 2018 web site was launched.

TUTORIAL: Multi-Objective Big Data Optimization with jMetalSP (JMETALSP)

By Antonio J. Nebro and Jose M. Garcia Nieto

Contents of the tutorial:

We are immersed in the Big Data era, where applications must manage and analyze huge amounts of data that cannot be processed with traditional technologies. The volume of data is not the only feature to take into account by these kinds of applications, since they also have to deal with heterogeneous data sources, many of them in streaming, and often delivering data at a great velocity. These data must furthermore be validated and analyzed to produce a significant value to the final user. These features comprise the so-called V's of Big Data: volume, velocity, variety, variability, veracity and value.

In this context, evolutionary algorithms and metaheuristics in general will play an important role in solving Big Data optimization problems. If we focus on multi-objective problems, i.e., those having two or more conflicting objectives, they can be found in many disciplines such as transportation, economics, medicine and bioinformatics, so Big Data variants of these problems will be common in the near future. Another important issue is that some real-world multi-objective problems involve objectives, constraints, and parameters that can change over time. These problems are referred to as dynamic multi-objective optimization problems, and they are related to Big Data optimization because they may change due to the analysis of streaming data which are received continuously from different data sources.

In this tutorial we describe the main features of jMetalSP, a Java-based framework designed for solving multi-objective Big Data optimization problems. jMetalSP combines the jMetal framework with Apache Spark, resulting in a tool that allows to easily use and adapt the metaheuristics provided by the former and take advantage of the distriting computing facilities of the latter. We present the architecture of jMetalSP and its main componentes, and show how multi-objective metaheuristics can be parallelized using Spark. Special attention will be parallelized to dynamic multi-objective problems and the analysis of streaming data sources. We use a transportation problem to illustrate the working of jMetalSP.

Intended audience:

Researchers and practitioners interested in Big Data applicacions requiring optimization and streaming data processing.

Tutorial format:

Mainly practical. The material needed to follow the tutorial will be available in a public repository beforehand.

About the presenters:

Antonio J. Nebro received his M.S. and Ph.D. degrees in Computer Science in 1992 and 1999, respectively, from the University of Malaga, Spain. He is currently an Associate Professor of Computer Science in this university. His current research activity is related to multi-objective optimization algorithms and parallelism, software tools for multi-objective optimization, and the application of these techniques to real-world problems of different domains, including bioinformatics, civil engineering, and Big Data. He is one of the designers and the main developer of the jMetal and jMetalSP frameworks for multi-objective optimization with metaheuristics.

Jose M. Garcia Nieto is a Post-Doctoral researcher at the University of Malaga. He received his M.S. and Ph.D. degrees in Computer Science in 2007 and 2013, respectively, from the University of Malaga. His current research topics are optimization algorithms, with special interest on swarm intelligence and multi-objective optimization, and their application to real-parameter benchmarking and real-world problems in the domains of bioinformatics and smart cities. In the last years, he is intensifying his research activity towards data analytics on Big Data and streaming processing environments, as well as on Web Semantics and Linked Data techniques for knowledge acquisition.