I am looking at a number of algorithms in Data Mining and some are described as being greedy. My issue is that they seem to be using the term greedy in different ways, which seems contradictory.
For example, with Hierarchical Clustering the decision to merge clusters is described as greedy. However, a filter attribute selection method where we select a subset of the attributes which possess the highest Information Gain is also described as greedy because it does not cope well with the interaction of attributes.
What does the term greedy mean in terms of Data Mining?