I'm asked to learn a concept using the Version Spaces (VS) algorithm that is able to predict life expectancy given a set of samples and data. The class examples we saw are kind of trivial compared to the assignment we got now. In particular I'm unsure about how to decide on the granularity of the concept hierarchies given that the data is often highly fragmented.
Does anyone have an elaborate example of the VS algorithm, preferably with non-trivial data sets?