A Tight Analysis of Greedy Yields Subexponential Time Approximation for Uniform Decision Tree

Decision Tree is a classic formulation of active learning: given n hypotheses with nonnegative weights summing to 1 and a set of tests that each partition the hypotheses, output a decision tree using the provided tests that uniquely identifies each hypothesis and has minimum (weighted) average depth. Previous works showed that the greedy algorithm achieves a O(log n) approximation ratio for this problem and it is NP-hard beat a O(log n) approximation, settling the complexity of the problem. However, for Uniform Decision Tree, i.e. Decision Tree with uniform weights, the story is more subtle. The greedy algorithm's O(log n) approximation ratio was the best known, but the largest approximation ratio known to be NP-hard is 4-varepsilon. We prove that the greedy algorithm gives a O(fraclog nlog COPT) approximation for Uniform Decision Tree, where COPT is the cost of the optimal tree and show this is best possible for the greedy algorithm. As a corollary, we resolve a conjecture of Kosaraju, Przytycka, and Borgstrom. Leveraging this result, for all α∈(0,1), we exhibit a frac9.01α approximation algorithm to Uniform Decision Tree running in subexponential time 2tilde O(nα). As a corollary, achieving any super-constant approximation ratio on Uniform Decision Tree is not NP-hard, assuming the Exponential Time Hypothesis. This work therefore adds approximating Uniform Decision Tree to a small list of natural problems that have subexponential time algorithms but no known polynomial time algorithms. All our results hold for Decision Tree with weights not too far from uniform. A key technical contribution of our work is showing a connection between greedy algorithms for Uniform Decision Tree and for Min Sum Set Cover.