Theoretical Analysis of Submodular Information Measures for Targeted Data Subset Selection
This work provides foundational theoretical support for a widely used method in machine learning data selection, addressing a key gap for researchers and practitioners.
The paper tackles the lack of theoretical guarantees for Submodular Mutual Information (SMI) in targeted data subset selection by deriving similarity-based bounds on relevance and coverage, showing that SMI functions are theoretically sound for achieving good query relevance and coverage.
With increasing volume of data being used across machine learning tasks, the capability to target specific subsets of data becomes more important. To aid in this capability, the recently proposed Submodular Mutual Information (SMI) has been effectively applied across numerous tasks in literature to perform targeted subset selection with the aid of a exemplar query set. However, all such works are deficient in providing theoretical guarantees for SMI in terms of its sensitivity to a subset's relevance and coverage of the targeted data. For the first time, we provide such guarantees by deriving similarity-based bounds on quantities related to relevance and coverage of the targeted data. With these bounds, we show that the SMI functions, which have empirically shown success in multiple applications, are theoretically sound in achieving good query relevance and query coverage.