Comparison-Based Random Forests
This addresses a problem for machine learning practitioners dealing with data where only relative comparisons are accessible, offering a novel approach that is not incremental but builds on random forests in a new context.
The paper tackles the problem of learning from data when only triplet comparisons (closer to B or C?) are available, without direct access to distances or representations, by proposing a random forest algorithm for regression and classification. The result shows that the method is efficient and competitive with approaches that use full metric information, as demonstrated in comprehensive experiments.
Assume we are given a set of items from a general metric space, but we neither have access to the representation of the data nor to the distances between data points. Instead, suppose that we can actively choose a triplet of items (A,B,C) and ask an oracle whether item A is closer to item B or to item C. In this paper, we propose a novel random forest algorithm for regression and classification that relies only on such triplet comparisons. In the theory part of this paper, we establish sufficient conditions for the consistency of such a forest. In a set of comprehensive experiments, we then demonstrate that the proposed random forest is efficient both for classification and regression. In particular, it is even competitive with other methods that have direct access to the metric representation of the data.