ADReM

Symposium on Pattern Discovery

Your are cordially invited to the Symposium on Pattern Discovery, featuring several invited talks on various data mining topics. The symposium will be held on Friday the 21st of October 2011. Afterwards it is followed by the public PhD defense of Michael Mampaey.

Programme:

  • 10:00h Welcome coffee
  • 10:30h Toon Calders (Eindhoven University of Technology): "Online Discovery of Top-k Similar Motifs in Time Series Data"
  • 11:15h Arno Siebes (Utrecht University): "Association Rules That Compress"
  • 12:00h Lunch break
  • 13:30h Geoff Webb (Monash University): "Finding Interesting Itemsets"
  • 14:15h Floris Geerts (University of Antwerp): "Data Quality: Research Opportunities in Data Mining"
  • 15:00h Coffee break
  • 16:00h Public PhD defense of Michael Mampaey: "Summarizing Data with Informative Patterns"

If you wish to attend, please go to http://adrem.ua.ac.be/michael.mampaey/phd for more information.

Event date and time: 
October 21, 2011 - 10:00 - 16:00

PhD Defense Michael Mampaey: "Summarizing Data with Informative Patterns"

Dissertation Michael Mampaey
Public defense of the PhD thesis entitled "Summarizing Data with Informative Patterns", by Michael Mampaey.

The defense is public and takes place at: Promotiezaal, klooster van de Grauwzusters, Lange Sint-Annastraat 7, 2000 Antwerpen. For more information, see http://adrem.ua.ac.be/michael.mampaey/phd.

Event date and time: 
October 21, 2011 - 16:00 - 18:00

Guest speaker: Prof. Szymon Jaroszewicz - Decision trees for uplift modeling

On Wednesday april 6, guest professor Szymon Jaroszewicz from the Institute of Computer Science of the Polish Academy of Sciences, and the National Institute of Telecommunications in Poland, will give an invited presentation on "Decision trees for uplift modeling" at 4pm in room G.006.

Abstract:
Traditional classification models predict the probability that an object belongs to a given class. In many cases, such as clinical trials or direct marketing, some action is to be taken based on model's predictions. In such a case one wants, instead, to model the difference between class probabilities in the case when the action is taken and the case when the action is not taken. Such a model, called an 'uplift model' makes it possible to find individuals for which the action is most profitable. This talk will give an overview of uplift modeling and present author's recent contributions to building uplift decision trees.

Event date and time: 
April 6, 2011 - 16:00 - 17:00

Guest lecture on Nature-Inspired Autonomous Systems by Karl Tuyls

Karl TuylsKarl TuylsOn monday March 21st, Prof. Karl Tuyls from Maastricht University will give an invited lecture in the Artificial Intelligence course. Karl Tuyls is head of Swarmlab, the robotics laboratory at the Department of Knowledge Engineering (DKE), Maastricht University. His main research interests lie at the intersection of Reinforcement Learning, Multi-Agent Systems and (Evolutionary) Game Theory. Everyone is invited to attend this guest lecture.

Title: "Nature-Inspired Autonomous Systems"

Abstract:
In this talk I will focus on how to build adaptive systems (agents or robots) capable of learning from their environment and peers in order to solve complex tasks. For this purpose we draw inspirations from nature and investigate bio-inspired techniques such as swarm intelligence (social insect behavior as found in honyebees and ants), reinforcement learning and evolutionary algorithms. I will discuss two research lines of the swarmlab research laboratory, i.e., one, how to design large-scale distributed, self-organizing robot systems capable of solving complex coordination tasks, such as foraging for food. These tasks find direct application in areas as automated patrolling, localization of danger and localization of victims in rescue management scenarios. Secondly, how to (optimally) learn from the environment and peers using a combination of reinforcement learning and evolutionary game theory in multi-agent systems in both cooperative and competitive situations. Finally, I will draw connections between both research lines.

Location: G004 (CMI)

Event date and time: 
March 21, 2011 - 10:45 - 12:45

Phd-Defense Wim Le Page: "Mining Patterns in Relational Databases"

Public defense of the PhD-thesis: “Mining Patterns in Relational Databases”, by Wim Le Page.

The defense is public and takes place in aula Jan Fabre (G0.10) of building G, Middelheimlaan 1, 2020 Antwerpen.

Abstract:

The Information Age has provided us with huge data repositories which cannot longer be analysed manually. The potential high business value of the knowledge that can be gained, drives the research for automated analysis methods that can handle large amounts of data. Most of the data of industry, education and government is stored in relational database management systems (RDBMSs). This motivates the need for data mining algorithms that can work with arbitrary relational
databases, without the need for manual transformation and preprocessing of the data. In this dissertation we introduce two data mining approaches towards the goal of mining interesting patterns in arbitrary relational databases.

Event date and time: 
December 16, 2009 - 16:00 - 18:00

PhD-Defense Adriana Prado: "An Inductive Database System Based on Virtual Mining Views"

Public defense of the PhD-thesis: "An Inductive Database System Based on Virtual Mining Views", by Adriana Prado.

The defense is public and takes place in aula Jan Fabre (G0.10) of building G, Middelheimlaan 1, 2020 Antwerpen.

Abstract: Data mining is an interactive process in which different tasks may be performed sequentially; the output of different tasks may be combined to be used as input for subsequent ones. In order to effectively support this knowledge discovery process, the integration of data mining into database systems has become necessary. The concept of inductive database systems has been proposed so as to achieve this integration. Contrary to the numerous proposals of data mining query languages, in this thesis, we present an inductive database system in which the query language is standard SQL. We propose a system in which the user can query the collection of all possible patterns as if they were stored in traditional relational tables.

Event date and time: 
December 14, 2009 - 16:00 - 18:00
Syndicate content