Develop a Machine Learning Life Cycle in Oracle Accelerated Data Science (ADS) SDK

This paper initially exposed the advantages of the Oracle Accelerated Data Science (ADS) SDK in the larger context of OCI features. The focus was on explaining the configuration steps for a machine learning life cycle. The most important steps related to environment setup and clarifying OCI and data science concepts were highlighted. To test the features of ADS SDK, a public dataset on higher education students’ performance evaluation with 33 attributes and 145 rows was used. I have described and explained the most important steps for a machine learning pipeline with ADS SDK, completing with the features selected, based on machine learning explainability computing. The advantage of ADS SDK, Conda environments, can be further exploited in my next machine learning experiments. To cross-validate the results provided by ADS SDK, I created an automated machine learning pipeline in Microsoft Automated ML, presented the results, and commented on the comparison between both technologies.