Learning Low-Dimensional Metrics
This work addresses theoretical gaps in metric learning for researchers, offering foundational insights but is incremental in nature.
The paper tackles the theoretical foundations of metric learning by developing generalization error bounds and sample complexity analysis for low-dimensional metrics, providing novel mathematical approaches and insights into ordinal embedding.
This paper investigates the theoretical foundations of metric learning, focused on three key questions that are not fully addressed in prior work: 1) we consider learning general low-dimensional (low-rank) metrics as well as sparse metrics; 2) we develop upper and lower (minimax)bounds on the generalization error; 3) we quantify the sample complexity of metric learning in terms of the dimension of the feature space and the dimension/rank of the underlying metric;4) we also bound the accuracy of the learned metric relative to the underlying true generative metric. All the results involve novel mathematical approaches to the metric learning problem, and lso shed new light on the special case of ordinal embedding (aka non-metric multidimensional scaling).