This Case study presents an example of applying COSMIC to measure the requirements of generic software that supports the training of Image Classifier software using a Machine Learning algorithm. This document segregates the functionality of the generic (‘classical’) software from the functionality specific to Machine Learning software, which will facilitate delegating tasks to staff with programming expertise, thereby freeing up time of Machine Learning data analysts. It also enables to collect data in a standardized fashion to develop estimation models for planning purposes and for on-going monitoring of the software tasks within a Machine Learning development project.
You can find it in the Knowledge base of cosmic-sizing, or download it here.