Exploring the Impact of BudgetPrune on Apache Spark Random Forest Performance
Keywords
Loading...
Authors
Issue Date
2017-06-29
Language
en
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This thesis explores the impact of a previously proposed pruning algorithm for random forest ensembles called BudgetPrune. BudgetPrune
tries to optimize the tradeoff between prediction accuracy and feature
acquisition cost, allowing for accurate prediction in resource-constrained
environments. Using Apache Spark ML's random forest model as a baseline, the influence of the pruning step on prediction accuracy and cost is
examined.
Description
Citation
Supervisor
Faculty
Faculteit der Sociale Wetenschappen