The Parameterized Complexity of Clustering Incomplete Data

We study fundamental clustering problems for incomplete data. Specifically, given a set of d-dimensional vectors (representing rows matrix), the goal is to complete missing vector entries in way that admits partitioning into at most k clusters with radius or diameter r. give tight characterizations parameterized complexity these respect parameters k, r, and minimum number columns needed cover all entries. show considered are fixed-parameter tractable when by three combined, dropping any results intractability. A byproduct our that, data setting, under consideration k+r.