Barranquilla, Colombia

June 5th ‐ 7th

TUTORIALS


Tutorial 1: Deep Learning 

This tutorial will deal with an introduction to artificial neural networks (ANN) and review the advantages of the current frameworks to design networks with multiple layers. Differences between sequential and functional of ANN models based on the keras API and its use for establishing multimodal networks. In addition, basic concepts related with architectures of ANN as batch normalization, dropout and transfer learning will be exposed. For this, different tools provided by Tensorflow will be analyzed for ANN training. Finally, examples with diverse architectures will be shown in applications for image, text and time series processing.

 

 

 

Julián D. Arias Londoño

received the B.S. degree in electronic engineering and MEng. degree in engineering – industrial automation from the Universidad Nacional de Colombia, Manizales, Colombia, in 2005 and 2007, respectively. He also received a Ph.D degree in Computer Science and Automatics jointly between the Universidad Politécnica de Madrid, Spain and Universidad Nacional de Colombia in 2010. Julián is Associate Professor of the Department of Systems Engineering and Computer Science, Universidad de Antioquia, Medellín, Colombia and part of the Intelligence information Systems Lab at the same University. His fields of expertise are in the areas of computational intelligence, machine learning and signal processing, applied to the design and development of computer-aided diagnostic system and Machine Learning – based solutions. Julián is member of INSTICC and Senior member of the Institute of Electrical and Electronics Engineers (IEEE) and its Signal Processing and Computational Intelligent societies. He has been leader of different research and development projects, which were funded by public and private institutions, and with participation of Colombian companies.


Turorial 2:  Programación y Técnicas de Estimación de Canal

La capacidad de comunicarse es la principal razón por la cual el hombre ha alcanzado sus niveles de desarrollo y evolución. Esa capacidad para transmitir información permite que sus ideas sean realizadas por un grupo de personas o una comunidad. A medida que hemos incrementado en número la población global se ha requerido de medios que permitan transmitir información a mayor número de personas, distancia y eficiencia.  En este sentido, la eficiencia puede interpretarse como el aumento de la velocidad, aumento de la cantidad de información sin perder la calidad.

 

 

 

 

Jorge Gómez Rojas

is a Tenured Professor in signal processing and Radiocommunication Systems at the Faculty of Engineering, at University of  Magdalena. His research interests centre around Physical layer (PHY), processing signal and radio propagation. He has experience of working in several international enterprises such as Impsat Fiber Networks, Siemens, and Ecopetrol.

He completed a Bachelor of Electronic Engineering (1999), Master in Electronic Engineering (2008) and a PhD in Engineering from Pontificia Bolivariana University (2018). from 2014, He is a Senior Member IEEE. His research interests lie primarily in applications of  Game Engine using Ray Launching technique to predict waves propagations. In addition, he has published works (books and papers) in propagation area.

 


Turorial 3:  Evolutionary Computing

In the tutorial we will describe major techniques of evolutionary computing that represent nature-inspired robust optimization methods tailored to various problem domains. We will introduce traditional genetic algorithms working on discrete domains, evolutionary strategies as a powerful continuous optimization framework, and genetic programming evolving tree-like structures representing computer programs. Several applications of these techniques to solve problems from optimization and machine learning areas will be demonstrated. Throughout the tutorial we will show concrete practical examples utilizing the Python programming language with optimization and machine learning tools.

 

 

 

 

Raman Neruda

Roman Neruda is with the Institute of Computer Science of the Czech Academy of Sciences, department of machine learning, where he is working in the areas of neurocomputing, evolutionary algorithms and meta-learning. He graduated from the Faculty of Mathematics and Physics, Charles University, and obtained his CSc degree in the ICS CAS. In 1995-1996 he was with the Los Alamos National Laboratory, he has been working on joint project with colleagues from   Carnegie-Mellon University, Koblenz Universitaet, University of California Chico, University of St. Etienne, and Universidad Distrital Bogota. He is the co-author of more than a hundred international publications. He teaches evolutionary algorithms and multi-agent systems at Faculty of Mathematics and Physics, Charles University