QoS-based data analytic service selection: a comparative study of different learning models

Abstract

QoS-based web service selection has been studied in the service computing community for some time; however, data characteristics are not considered. In this work, we have studied the use of different machine learning algorithms as meta-learners in predicting the performance of data analytic services for the given dataset. We used a meta-learning algorithm to incorporate meta-features in the selection process and we used clustering services as an example of data analytic services. We have also investigated the impact of the number of data features on the performance of the meta-learners. We found that, out of the 5 classification models, SVM showed the best results in predicting the recommended service for the given dataset with an accuracy of 78%. When it comes to regression models, MLP was the best regressor. We recommend considering only simple meta-features that can be collected for most datasets, as those proved to be sufficient to achieve good prediction accuracy.