Course: UoIceland Teaching

Project Details

Project Lead
Morris Riedel 
Project Manager
Morris Riedel 
Project Members
Thomas Philip Runarsson, Jeff Hair, Camille Bernard, Leo Rabel, Hannes Eggertsson, Atli Ómarsson, Alessandro Cremona, Bjorn Thorgilsson, Maximilian Voit, Sigurdur Hannesson, Alexandre Legrain, Dren Kajmakci, Gabriele Cavallaro, Ingvi Stígsson, Tryggvi Thorsteinsson, Johannes Agustarson, Matthías Ásgeirsson, Kristmundur Ólafsson, Jon Oli Olafsson, Stephen Onyango, Einar Indridason, Oscar Cideos, Magnus Gislason, Tom Schiller, Stefan Thorvardarson, Bodvar Sveinsson, Sindri Bjarnason, Björn Hagemeier, Philipp Glock, Steinunn Groa Siguroardottir, Inga Rún Helgadóttir  
Institution
Juelich Supercomputing Centre, Federated Systems and Data  
Discipline
Computer Science (401) 
Subdiscipline
30.08 Mathematics and Computer Science 

Abstract

The course is part of teaching courses of the University of Iceland. It aims to provide practical insights into some data mining & machine learning approaches/algorithms (e.g. Apache Mahout [1]) in combination with (batch) execution frameworks (e.g. HPC, Apache Hadoop/HTC [2], Twister [3], etc.). I could offer practical insights into some data mining&machine learning approaches/algorithms (e.g. Apache Mahout [1]) in combination with (batch) execution frameworks (e.g. HPC, Apache Hadoop/HTC [2], Twister [3], etc.).


 

Intellectual Merit

In order to understand the theoretical approaches of statistical data mining & machine learning the students require to learn hands-on experience with real data and real frameworks.

Broader Impacts

Findings of student projects on the resources will be brought into the Research Data Alliance (RDA) Big Data Analytics Group that is chaired by this project coordinator. FutureGrid will be references and cited in each of the findings we present there.

Scale of Use

For a couple of weeks for a class (~20 students)

Results

Students will have an understanding of the open source frameworks in the field of data mining, machine learning in combination with (batch) execution frameworks. Selected contributions of student project results will be given as an input to the RDA Big Data Analytics Group that in turn creates a classification of feasible big data analytics approaches including algorithms, frameworks, and underlying resources.