computational and statistical tradeoff
Computational and Statistical Tradeoffs in Learning to Rank
For massive and heterogeneous modern data sets, it is of fundamental interest to provide guarantees on the accuracy of estimation when computational resources are limited. In the application of learning to rank, we provide a hierarchy of rank-breaking mechanisms ordered by the complexity in thus generated sketch of the data. This allows the number of data points collected to be gracefully traded off against computational resources available, while guaranteeing the desired level of accuracy. Theoretical guarantees on the proposed generalized rank-breaking implicitly provide such trade-offs, which can be explicitly characterized under certain canonical scenarios on the structure of the data.
Computational and Statistical Tradeoffs in Learning to Rank
Neural Information Processing Systems http://nips.cc/
Reviews: Computational and Statistical Tradeoffs in Learning to Rank
Novelty: - I understand the method, but I'm just a bit surprised that it does better (empirically) than using pairwise comparisons in an "intelligent" way, i.e., [3, 15]. Can you explain why? - Actually I'm a bit confused here. You write that [3, 15] are consistent (L94, 100), but write in your legends "inconsistent PRB" (Figs. 2, 3), and show that these methods behave inconsistently in these plots. Can you clarify? - Also, I wonder if your method really worth it? How long does "inconsistent PRB" take to run (we know your method's runtime from Figure 1 left)?
Computational and Statistical Tradeoffs in Learning to Rank
For massive and heterogeneous modern data sets, it is of fundamental interest to provide guarantees on the accuracy of estimation when computational resources are limited. In the application of learning to rank, we provide a hierarchy of rank-breaking mechanisms ordered by the complexity in thus generated sketch of the data. This allows the number of data points collected to be gracefully traded off against computational resources available, while guaranteeing the desired level of accuracy. Theoretical guarantees on the proposed generalized rank-breaking implicitly provide such trade-offs, which can be explicitly characterized under certain canonical scenarios on the structure of the data. Papers published at the Neural Information Processing Systems Conference.