MOAManager: A tool to support data stream experiments. (29th January 2020)

• Record Type: Journal Article
• Authors: Maciel, Bruno I. F.; Santos, Silas G. T. C.; Barros, Roberto S. M.
• Is Part Of: Software, practice & experience. Volume 50:Number 4(2020)
• Page Start: 325
• Page End: 334
• Publication Date: 2020-01-29
• Subjects: concept drift -- online learning -- performance analysis -- PHP -- statistical analysis -- tool; Computer software -- Periodicals; Computer programming -- Periodicals; Computer programs -- Periodicals; 005.3
• DOI: 10.1002/spe.2795
• Languages: English
• ISSNs: 0038-0644

Abstract

Summary: This short communication describes Massive Online Analysis Manager Hypertext Preprocessor framework (MOAManager), an open‐source web tool that assists in the creation, execution, extraction, and formatting of experiments in online environments using the MOA framework. MOAManager also brings together a set of features for data analysis, including the calculation of metrics and statistical tests for performance evaluation. The purpose of the tool is to provide a structure that automates several repetitive experimental steps, helping the users to customize, run, and analyze the experiments quickly and reliably, enabling gains in productivity.