Abstract
In this paper we initiate study of the computational power of adaptive and non-adaptive monotone decision trees - decision trees where each query is a monotone function on the input bits. In the most general setting, the monotone decision tree height (or size) can be viewed as a measure of non-monotonicity of a given Boolean function. We also study the restriction of the model by restricting (in terms of circuit complexity) the monotone functions that can be queried at each node. This naturally leads to complexity classes of the form \({\mathsf {DT}}(\textit{mon-}\mathcal {C})\) for any circuit complexity class \(\mathcal {C}\), where the height of the tree is \(\mathcal {O}(\log n)\), and the query functions can be computed by monotone circuits in class \(\mathcal {C}\). In the above context, we prove the following characterizations and bounds.
-
We show that the decision tree height can be exactly characterized (both in the adaptive and non-adaptive versions of the model) in terms of the alternation (\({\mathsf {alt}}(f)\)) of a function (defined as the maximum number of times that the function value changes, in any chain in the Boolean lattice). We also characterize the non-adaptive decision tree height with a natural generalization of certification complexity of a function. We also show upper bounds and characterizations for non-deterministic and randomized variants of the monotone decision trees in terms of \({\mathsf {alt}}(f)\).
-
We show that \({\mathsf {DT}}(\textit{mon-}\mathcal {C}) = \mathcal {C}\) when \(\mathcal {C}\) contains monotone circuits for the threshold functions. For \(\textsf {AC}^0\), we show that any function in \(\textsf {AC}^0\) can be computed by a sub-linear height monotone decision tree with queries having monotone \(\textsf {AC}^0\) circuits.
-
To explore the logarithmic height case - \({\mathsf {DT}}(\textit{mon-}\textsf {AC}^0)\) - we show that for any f (on n bits) in \({\mathsf {DT}}(\textit{mon-}\textsf {AC}^0)\), and for any constant \(0<\epsilon \le 1\), there is an \(\textsf {AC}^0\) circuit for f with \(\mathcal {O}(n^\epsilon )\) negation gates. In contrast, it can be derived from [14] that for every \(f \in \textsf {AC}^0\) with \({\mathsf {alt}}(f) = \varOmega (n)\), and for every \(\epsilon > 0\), any \(\textsf {AC}^0\) circuit computing it with \(\mathcal {O}(n^\epsilon )\) negations will need at least \(\frac{1}{\epsilon }\) depth.
En route the main results, as a tool, we study the monotone variant of the decision list model, and prove corresponding characterizations in terms of \({\mathsf {alt}}(f)\) and also derive as a consequence that \({\mathsf {DT}}(\textit{mon-}\mathcal {C}) = {\mathsf {DL}}(\textit{mon-}\mathcal {C})\) if \(\mathcal {C}\) has appropriate closure properties (where \({\mathsf {DL}}(\textit{mon-}\mathcal {C})\) is defined similar to \({\mathsf {DT}}(\textit{mon-}\mathcal {C})\) but for decision lists).
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We denote the set \(\{1,2,\dots n\}\) by [n].
- 2.
If \(\mathsf{supp}(f) = \{S \subseteq [n] \mid \hat{f}(S) \ne 0\}\), \(\mathsf{sps}(f) = |\mathsf{supp}(f)|\) and \(\mathsf{fdim}(f) = \dim (\mathsf{supp}(f))\), then by [19], \(\log \mathsf {sps}(f)/2 \le \mathsf {DT_{\oplus }}(f) \le \mathsf {fdim}(f) = \mathsf {DT}_{\oplus }^\mathsf{na}(f)\) [9, 15]. The XOR-logrank conjecture [19] states that \(\mathsf {DT_{\oplus }}(f) \le \textsf {poly}\left( \log \mathsf {sps}(f)\right) \), and \(\exists f\) for which \(\mathsf {fdim}(f)\) and \(\log (\mathsf{sps}(f))\) are exponentially far apart.
- 3.
Indeed, even though the queries are restricted to monotone functions on inputs, the model is universal, since in normal decision trees, the queries are already monotone.
- 4.
It is assumed that in \(\mathcal {C}\), all the circuits are polynomial sized, and that there is atleast one function with \(\varOmega (n)\) alternation.
- 5.
Using the same constructions, we also observe that \({\mathsf {DT}}(\textit{mon-}\mathcal {C})={\mathsf {DL}}(\textit{mon-}\mathcal {C})\) for any circuit complexity class \(\mathcal {C}\) with appropriate closure properties.
- 6.
In the full version of this paper, we define another natural variant of non-determinism, and show its equivalence to this model.
References
Ben-Asher, Y., Newman, I.: Decision trees with and or queries. In Proceedings of 10th Annual Conference on Structure in Complexity Theory, pp. 74–81 (1995)
Ben-Or, M.: Lower bounds for algebraic computation trees. In: Proceedings of the 15th ACM Symposium on Theory of Computing, pp. 80–86 (1983)
Björner, A., Lovász, L., Yao, A.C.C.: Linear decision trees: volume estimates and topological bounds. In Proceedings of the 24th ACM Symposium on Theory of Computing, pp. 170–177 (1992)
Blais, E., Canonne, C.L., Oliveira, I.C., Servedio, R.A., Tan, L.-Y.: Learning circuits with few negations. Approx. Random. Comb. Opti. 40, 512–527 (2015)
Blum, A.: Rank-r decision trees are a subclass of r-decision lists. Inf. Process. Lett. 42(4), 183–185 (1992)
Bshouty, N.H.: A subexponential exact learning algorithm for DNF using equivalence queries. Inf. Process. Lett. 59, 37–39 (1996)
Buhrman, H., de Wolf, R.: Complexity measures and decision tree complexity: a survey. Theor. Comput. Sci. 288(1), 21–43 (2002)
Dobkin, D., Lipton, R.J.: A lower bound of n on linear search programs for the knapsack problem. J. Comput. Syst. Sci. 16, 413–417 (1978)
Gopalan, P., O’Donnell, R., Servedio, R.A., Shpilka, A., Wimmer, K.: Testing Fourier dimensionality and sparsity. SIAM J. Comput. 40(4), 1075–1100 (2011)
Guijarro, D., Lavin, V., Raghavan, V.: Monotone term decision lists. Theoret. Comput. Sci. 259(1), 549–575 (2001)
Jukna, S.: Boolean Function Complexity: Advances and Frontiers. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-24508-4
Kushilevitz, E., Mansour, Y.: Learning decision trees using the Fourier spectrum. SIAM J. Comput. 22(6), 1331–1348 (1993)
Markov, A.A.: On the inversion complexity of a system of functions. J. ACM 5(4), 331–334 (1958)
Santha, M., Wilson, C.: Limiting negations in constant depth circuits. SIAM J. Comput. 22(2), 294–302 (1993)
Sanyal, S.: Fourier sparsity and dimension. Theory Comput. 15(11), 1–13 (2019)
Snir, M.: Proving lower bounds for linear decision trees. In: Even, S., Kariv, O. (eds.) ICALP 1981. LNCS, vol. 115, pp. 305–315. Springer, Heidelberg (1981). https://doi.org/10.1007/3-540-10843-2_25
Steele, J.M., Yao, A.C.: Lower bounds for algebraic decision trees. Technical report, Department of Computer Science, Stanford University (1980)
Yao, A.C., Rivest, R.L.: On the polyhedral decision problem. SIAM J. Comput. 9(2), 343–347 (1980)
Zhang, Z., Shi, Y.: On the parity complexity measures of Boolean functions. Theoret. Comput. Sci. 411(26), 2612–2618 (2010)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Amireddy, P., Jayasurya, S., Sarma, J. (2020). Power of Decision Trees with Monotone Queries. In: Kim, D., Uma, R., Cai, Z., Lee, D. (eds) Computing and Combinatorics. COCOON 2020. Lecture Notes in Computer Science(), vol 12273. Springer, Cham. https://doi.org/10.1007/978-3-030-58150-3_23
Download citation
DOI: https://doi.org/10.1007/978-3-030-58150-3_23
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-58149-7
Online ISBN: 978-3-030-58150-3
eBook Packages: Computer ScienceComputer Science (R0)