More Like this

1. On the Power of Adaptivity for Function Inversion

   https://doi.org/10.4230/LIPIcs.ITC.2024.5  

   Gajulapalli, Karthik; Golovnev, Alexander; King, Samuel (January 2024, Schloss Dagstuhl – Leibniz-Zentrum für Informatik)

   Aggarwal, Divesh (Ed.)

   We study the problem of function inversion with preprocessing where, given a function f : [N] → [N] and a point y in its image, the goal is to find an x such that f(x) = y using at most T oracle queries to f and S bits of preprocessed advice that depend on f. The seminal work of Corrigan-Gibbs and Kogan [TCC 2019] initiated a line of research that shows many exciting connections between the non-adaptive setting of this problem and other areas of theoretical computer science. Specifically, they introduced a very weak class of algorithms (strongly non-adaptive) where the points queried by the oracle depend only on the inversion point y, and are independent of the answers to the previous queries and the S bits of advice. They showed that proving even mild lower bounds on strongly non-adaptive algorithms for function inversion would imply a breakthrough result in circuit complexity. We prove that every strongly non-adaptive algorithm for function inversion (and even for its special case of permutation inversion) must have ST = Ω(N log (N) log (T)). This gives the first improvement to the long-standing lower bound of ST = Ω(N log N) due to Yao [STOC 90]. As a corollary, we conclude the first separation between strongly non-adaptive and adaptive algorithms for permutation inversion, where the adaptive algorithm by Hellman [TOIT 80] achieves the trade-off ST = O(N log N). Additionally, we show equivalence between lower bounds for strongly non-adaptive data structures and the one-way communication complexity of certain partial functions. As an example, we recover our lower bound on function inversion in the communication complexity framework. 

   more » « less
2. The (Im)possibility of Simple Search-To-Decision Reductions for Approximation Problems

   https://doi.org/10.4230/LIPIcs.APPROX/RANDOM.2023.10  

   Golovnev, Alexander; Guo, Siyao; Peters, Spencer; Stephens-Davidowitz, Noah (January 2023, Schloss Dagstuhl – Leibniz-Zentrum für Informatik)

   Megow, Nicole; Smith, Adam (Ed.)

   We study the question of when an approximate search optimization problem is harder than the associated decision problem. Specifically, we study a natural and quite general model of black-box search-to-decision reductions, which we call branch-and-bound reductions (in analogy with branch-and-bound algorithms). In this model, an algorithm attempts to minimize (or maximize) a function f: D → ℝ_{≥ 0} by making oracle queries to h_f : T → ℝ_{≥ 0} satisfying min_{x ∈ S} f(x) ≤ h_f(S) ≤ γ ⋅ min_{x ∈ S} f(x) (*) for some γ ≥ 1 and any subset S in some allowed class of subsets T; of the domain D. (When the goal is to maximize f, h_f instead yields an approximation to the maximal value of f over S.) We show tight upper and lower bounds on the number of queries q needed to find even a \gamma'-approximate minimizer (or maximizer) for quite large \gamma'; in a number of interesting settings, as follows. - For arbitrary functions f : {0,1}ⁿ → ℝ_{≥ 0}, where T; contains all subsets of the domain, we show that no branch-and-bound reduction can achieve γ' ≲ γ^{n/log q}, while a simple greedy approach achieves essentially γ^{n/log q}. - For a large class of MAX-CSPs, where T = {S_w} contains each set of assignments to the variables induced by a partial assignment w, we show that no branch-and-bound reduction can do significantly better than essentially a random guess, even when the oracle h_f guarantees an approximation factor of γ ≈ 1+√{log(q)/n}. - For the Traveling Salesperson Problem (TSP), where T = {S_p} contains each set of tours extending a path p, we show that no branch-and-bound reduction can achieve γ' ≲ (γ-1) n/log q. We also prove a nearly matching upper bound in our model. These results show an oracle model in which approximate search and decision are strongly separated. (In particular, our result for TSP can be viewed as a negative answer to a question posed by Bellare and Goldwasser (SIAM J. Comput. 1994), though only in an oracle model.) We also note two alternative interpretations of our results. First, if we view h_f as a data structure, then our results unconditionally rule out black-box search-to-decision reductions for certain data structure problems. Second, if we view h_f as an efficiently computable heuristic, then our results show that any reasonably efficient branch-and-bound algorithm requires more guarantees from its heuristic than simply Eq. (*). Behind our results is a ``useless oracle lemma'' which allows us to argue that under certain conditions the oracle h_f is ``useless'' and which might be of independent interest. 

   more » « less
3. The Non-Uniform Perebor Conjecture for Time-Bounded Kolmogorov Complexity Is False

   https://doi.org/10.4230/LIPIcs.ITCS.2024.80  

   Mazor, Noam; Pass, Rafael (January 2024, Schloss Dagstuhl – Leibniz-Zentrum für Informatik)

   Guruswami, Venkatesan (Ed.)

   The Perebor (Russian for "brute-force search") conjectures, which date back to the 1950s and 1960s are some of the oldest conjectures in complexity theory. The conjectures are a stronger form of the NP ≠ P conjecture (which they predate) and state that for "meta-complexity" problems, such as the Time-bounded Kolmogorov complexity Problem, and the Minimum Circuit Size Problem, there are no better algorithms than brute force search. In this paper, we disprove the non-uniform version of the Perebor conjecture for the Time-Bounded Kolmogorov complexity problem. We demonstrate that for every polynomial t(⋅), there exists of a circuit of size 2^{4n/5+o(n)} that solves the t(⋅)-bounded Kolmogorov complexity problem on every instance. Our algorithm is black-box in the description of the Universal Turing Machine U employed in the definition of Kolmogorov Complexity and leverages the characterization of one-way functions through the hardness of the time-bounded Kolmogorov complexity problem of Liu and Pass (FOCS'20), and the time-space trade-off for one-way functions of Fiat and Naor (STOC'91). We additionally demonstrate that no such black-box algorithm can have circuit size smaller than 2^{n/2-o(n)}. Along the way (and of independent interest), we extend the result of Fiat and Naor and demonstrate that any efficiently computable function can be inverted (with probability 1) by a circuit of size 2^{4n/5+o(n)}; as far as we know, this yields the first formal proof that a non-trivial circuit can invert any efficient function. 

   more » « less
4. Efficient Statistics, in High Dimensions, from Truncated Samples

   Daskalakis, C.; Gouleakis, T.; Tzamos, C.; Zampetakis, M. (October 2018, Annual Symposium on Foundations of Computer Science)

   We provide an efficient algorithm for the classical problem, going back to Galton, Pearson, and Fisher, of estimating, with arbitrary accuracy the parameters of a multivariate normal distribution from truncated samples. Truncated samples from a d-variate normal N(μ,Σ) means a samples is only revealed if it falls in some subset S⊆Rd; otherwise the samples are hidden and their count in proportion to the revealed samples is also hidden. We show that the mean μ and covariance matrix Σ can be estimated with arbitrary accuracy in polynomial-time, as long as we have oracle access to S, and S has non-trivial measure under the unknown d-variate normal distribution. Additionally we show that without oracle access to S, any non-trivial estimation is impossible. 

   more » « less
5. A Theory of Selective Prediction

   Qiao, Mingda; Valiant, Gregory (January 2019, Conference on Learning Theory (COLT))

   We consider a model of selective prediction, where the prediction algorithm is given a data sequence in an online fashion and asked to predict a pre-specified statistic of the upcoming data points. The algorithm is allowed to choose when to make the prediction as well as the length of the prediction window, possibly depending on the observations so far. We prove that, even without any distributional assumption on the input data stream, a large family of statistics can be estimated to non-trivial accuracy. To give one concrete example, suppose that we are given access to an arbitrary binary sequence x1, . . . , xn of length n. Our goal is to accurately predict the average observation, and we are allowed to choose the window over which the prediction is made: for some t < n and m < n − t, after seeing t observations we predict the average of x_{t+1},..., x{t+m}. This particular problem was first studied in [Dru13] and referred to as the “density prediction game”. We show that the expected squared error of our prediction can be bounded by O(1/log n) and prove a matching lower bound, which resolves an open question raised in [Dru13]. This result holds for any sequence (that is not adaptive to when the prediction is made, or the predicted value), and the expectation of the error is with respect to the randomness of the prediction algorithm. Our results apply to more general statistics of a sequence of observations, and we highlight several open directions for future work. 

   more » « less