On the Rational Degree of Boolean Functions and Applications

Oct 11, 2023
17 pages
e-Print:

Citations per year

0 Citations
Abstract: (arXiv)
We study a natural complexity measure of Boolean functions known as the (exact) rational degree. For total functions ff, it is conjectured that rdeg(f)\mathrm{rdeg}(f) is polynomially related to deg(f)\mathrm{deg}(f), where deg(f)\mathrm{deg}(f) is the Fourier degree. Towards this conjecture, we show that symmetric functions have rational degree at least deg(f)/2\mathrm{deg}(f)/2 and monotone functions have rational degree at least deg(f)\sqrt{\mathrm{deg}(f)}. We observe that both of these lower bounds are tight. In addition, we show that all read-once depth-dd Boolean formulae have rational degree at least Ω(deg(f)1/d)\Omega(\mathrm{deg}(f)^{1/d}). Furthermore, we show that almost every Boolean function on nn variables has rational degree at least n/2O(n)n/2 - O(\sqrt{n}). In contrast to total functions, we exhibit partial functions that witness unbounded separations between rational and approximate degree, in both directions. As a consequence, we show that for quantum computers, post-selection and bounded-error are incomparable resources in the black-box model.
Note:
  • 17 pages, 3 figures