University of Konstanz
Graduiertenkolleg / PhD Program
Computer and Information Science

Graduation Talks

title

Parallel Data Mining Revisited - Getting Better, Not Faster

speaker

Zaenal Akbar, University Konstanz
Konstanz, Germany

date & place

Wednesday, 25.04.2012, 15:15 h
Room C252

abstract

Many parallelization methods for data mining have been introduced whereby the primary objective is to improve performance or handle more data. In addition to data mining performance (speed-up), another important as- pect is accuracy. In this work we would like to investigate a solution to improve the accuracy of a greedy heuristic, a widely used algorithm in data mining which leaves room for improvement by reducing its greediness. Index Terms—parallel, data mining, accuracy, greedy heuristic