Exercising Good Practices of Machine Learning Research: A Case Study of Environment Image Classification
Abstract: It is of the utmost importance that, in both research and industry applications, results in the field of Machine Learning are performed and presented in a fair, explainable, and reproducible fashion. This paper uses the framework of an image classification case study to explore the practical application of a range of fundamental approaches that can guide such effective practices. We discuss and implement ideas of data collection and analysis, fairness, evaluation metrics, statistical interpretation, model implementation, repeatability, as well as the encouragement and provision of necessary resources for future research and cross-checking.
Submission Length: Long submission (more than 12 pages of main content)
Assigned Action Editor: ~Mathurin_Massias1
Submission Number: 3142
Loading