We investigate the influences of variables on a booleanfunction f based on the quantum Bernstein-Vazirani algorithm. A previous paper (Floess et al. in Math Struct Comput Sci 23:386, 2013) has proved that if an n-var...
详细信息
We investigate the influences of variables on a booleanfunction f based on the quantum Bernstein-Vazirani algorithm. A previous paper (Floess et al. in Math Struct Comput Sci 23:386, 2013) has proved that if an n-variable booleanfunction f (x(1) ,..., x(n)) does not depend on an input variable x(i), using the Bernstein-Vazirani circuit for f will always output y that has a 0 in the ith position. We generalize this result and show that, after running this algorithm once, the probability of getting a 1 in each position i is equal to the dependence degree of f on the variable x(i), i.e., the influence of x(i) on f. Based on this, we give an approximation algorithm to evaluate the influence of any variable on a booleanfunction. Next, as an application, we use it to study the booleanfunctions with juntas and construct probabilistic quantum algorithms to learn certain booleanfunctions. Compared with the deterministic algorithms given by Floess et al., our probabilistic algorithms are faster.
暂无评论