| Summary: | We wanted to determine whether we could use an automated machine learning system called Azure for the selection process and placement of conscript training in such a way that AI can make decisions for the right conscript training program individually. To test this, we had four separate datasets and access to the Microsoft Azure automated machine learning environment. According to the test sets we performed, we see that, by using an automated machine learning environment, it was possible to reach the precision level of the decisions we wanted. The main obstacle was not the used automated machine learning environment itself, but the quality of the data used for learning. We also made improvement suggestions regarding how data could be collected and what kind of data we should measure to make predictive data analysis better and be more usable in the future.
|
|---|