Average-case deterministic query complexity of boolean functions with fixed weight

Yuan Li,Haowei Wu,Yi Yang

2024-07-23

Abstract:We explore the $\textit{average-case deterministic query complexity}$ of boolean functions under a $\textit{uniform distribution}$, denoted by $\mathrm{D}_\mathrm{ave}(f)$, the minimum average depth of zero-error decision tree computing a boolean function $f$. This measure has found several applications across diverse fields, yet its understanding is limited. We study $\mathrm{D}_\mathrm{ave}(f)$ of several functions, including the penalty shoot-out function, symmetric functions, linear threshold functions and the tribes functions. We prove $\mathrm{D}_\mathrm{ave}(f) \le \max \{ \log \frac{\mathrm{wt}(f)}{\log n} + O(\log \log \frac{\mathrm{wt}(f)}{\log n}), O(1) \}$ for every $n$-variable boolean function $f$, where $\mathrm{wt}(f)$ denotes the weight (the number of inputs on which $f$ outputs $1$). For any $4\log n \le m(n) \le 2^{n-1}$, we prove the upper bound is tight up to an additive logarithmic term for almost all $n$-variable boolean functions with weight $\mathrm{wt}(f) = m(n)$. Using Håstad's switching lemma or Rossman's switching lemma [Comput. Complexity Conf. 137, 2019], one can derive $\mathrm{D}_\mathrm{ave}(f) \leq n(1 - \frac{1}{O(w)})$ or $\mathrm{D}_\mathrm{ave}(f) \le n(1 - \frac{1}{O(\log s)})$ for CNF/DNF formulas of width $w$ or size $s$, respectively. We show that, for any $w \ge \log n + \log \log n + 3$, there exists a DNF formula of width $w$ and size $\lceil 2^w / w \rceil$ such that $\mathrm{D}_\mathrm{ave}(f) = n (1 - \frac{\log n}{\Theta(w)})$. In other words, we show the criticality upper bounds $O(w)$ and $O(\log s)$ are tight up to a multiplicative $\log n$ factor, providing evidence on the tightness of the switching lemmas.

Computational Complexity

What problem does this paper attempt to address?

The problems that this paper attempts to solve mainly focus on the research of the Average - Case Deterministic Query Complexity of Boolean functions under the uniform distribution. Specifically, the paper aims to:

1. **Explore and understand** the Average - Case Deterministic Query Complexity \(D_{\text{ave}}(f)\) of Boolean functions under the uniform distribution, which is the minimum average depth of a zero - error decision tree computing the Boolean function \(f\). Although this metric has been applied in multiple fields, the understanding of it is still limited.

2. **Provide new upper and lower bounds**: The paper provides new upper and lower bounds for \(D_{\text{ave}}(f)\) of various Boolean functions (such as penalty - kick - out functions, symmetric functions, linear threshold functions and tribe functions), and proves that these bounds are tight in some cases.

3. **Study the relationship between Criticality and average - case query complexity**: By applying the Switching Lemma of Håstad and Rossman, the paper explores the relationship between Criticality and \(D_{\text{ave}}(f)\), and provides evidence that the upper bounds of these lemmas are tight, up to a logarithmic factor.

4. **Reveal the relationships between different complexity metrics**: The paper also studies the relationships between different complexity metrics (such as \(D_{\text{ave}}(f)\), \(R_{\text{ave}}(f)\), \(C_{\text{ave}}(f)\), \(bs_{\text{ave}}(f)\) and \(s_{\text{ave}}(f)\)), especially for balanced monotone functions, whether there are polynomial relationships between these metrics.

5. **Construct specific functions**: The paper constructs some specific Boolean functions, shows the performance of these functions under different complexity metrics, and thus verifies the correctness and validity of the theoretical results.

Through these studies, the paper not only deepens the understanding of the Average - Case Deterministic Query Complexity, but also provides new tools and methods for future research.