An exploratory study on machine learning frameworks
This document describes a preliminary study on computing frameworks and technologies, for the purpose of developing machine learning (ML) system applications. Several frameworks, application programming interfaces and programming libraries for ML algorithms have been developed in the last few years, in a relatively short period of time, making difficult a decision on which one to chose in a particular application. This study reviews some criteria and performs a preliminary evaluation of some of the most used ML technologies for developing system applications, with the purpose to guide and facilitate the decision on which of them to apply, given a particular application.
Mavridis, I. and Karatza, H. (2017). Performance evaluation of cloud-based log file anal- ysis with apache hadoop and apache spark. Journal ofSystems and Software, 125.
MICS (2019). Minist´erio da ind´ustria, com´ercio e servic¸o. industria 4.0. http://www. industria40.gov.br/. [Online; accessed 18-March-2019].
Shi, S., Wang, Q., Xu, P., and Chu, X. (2016). Benchmarking state-of-the-art deep learn- ing software tools. CoRR, abs/1608.07249.
Viademonte, S., de Souza, C., Carneiro, N., Junior, J. F., and Lyra, W. (2018). A computa- tional framework for railway incident analysis: from data mining to data visualization. In AMCIS2018, New Orleans, Louisiana US.