• Aucun résultat trouvé

On the Complexity of A/B Testing

N/A
N/A
Protected

Academic year: 2021

Partager "On the Complexity of A/B Testing"

Copied!
24
0
0

Texte intégral

Loading

Figure

Figure 1: Comparison of different information terms in Bernoulli bandit models.
Figure 2: Experimental results (descriptions in text; plots best viewed in color).

Références

Documents relatifs

Table 2 Strong-motion stations of the Swiss national network with on-line data acquisition operational at the end of 2012.. Code Station

In this paper exact and approximate methods are given for computing the distribution of quadratic forms in normal variables. In statistical applications the interest centres in

We proposed a meta-algorithm, called unified gap-based exploration (UGapE), that unifies the two settings of the best arm(s) identification problem in stochastic multi-armed

The new results, which close the above-mentioned gap between the lower bound and the upper bound, guarantee that no learning method, given the generative model of the MDP, can

the above-mentioned gap between the lower bound and the upper bound of RL, guarantee that no learning method, given a generative model of the MDP, can be significantly more

Here we prove an asymptotically optimal lower bound on the information complexity of the k-party disjointness function with the unique intersection promise, an important special case

We would also like to remark that every problem in MMSNP is a finite union of connected MMSNP sentences (see Section 3), and that connected MMSNP sentences describe

Then note that if these modifications are performed, then, in order to get a strong orientation of G H , for every representative pair {u, v}, we would need to have the path of length