We investigate the unprovability of NP$\not\subseteq$P/poly in various fragments of arithmetic. The unprovability is usually obtained by showing hardness of propositional formulas encoding superpolynomial circuit lower bounds. Firstly, we discuss few relevant techniques and known theorems. Namely, natural proofs, feasible interpolation, KPT theorem, iterability, gadget generators etc. Then we prove some original results. We show the unprovability of superpolynomial circuit lower bounds for systems admitting certain forms of feasible interpolation (modulo a hardness assumption) and for systems roughly described as tree-like Frege systems working with formulas using only a small fraction of variables of the statement that is supposed to be proved. These results are obtained by proving the hardness of the Nisan-Wigderson generators in corresponding proof systems.

Title: Complexity Theory in Feasible Mathematics Author: Ján Pich Department: Department of Algebra Supervisor: Prof. RNDr. Jan Krajíček, DrSc., MAE Abstract: We study the provability of statements and conjectures from Complex- ity Theory in Bounded Arithmetic. First, modulo a hardness assumption, we show that theories weaker in terms of provably total functions than Buss's theory S1 2 cannot prove nk -size circuit lower bounds for SAT formalized as a Σb 2-formula LB(SAT, nk ). In particular, the true universal first-order theory in the language containing names for all uniform NC1 algorithms denoted TNC1 does not prove LB(SAT, n4kc ) where k ≥ 1, c ≥ 2 unless each function f ∈ SIZE(nk ) can be approximated by formulas Fn of subexponential size 2O(n1/c) with subexponential advantage: Px∈{0,1}n [Fn(x) = f(x)] ≥ 1/2 + 1/2O(n1/c) . Unconditionally, V 0 does not prove quasipolynomial nlog n -size circuit lower bounds for SAT. Considering upper bounds, we prove the PCP theorem in Cook's theory PV1. This includes a formalization of the (n, d, λ)-graphs in PV1. A consequence of the result is that Extended Frege proof system admits p-size proofs of tautologies encoding the PCP theorem. Keywords: Circuit Lower Bounds, Bounded Arithmetic, The PCP theorem

In this thesis, we show results for some well-studied problems from learning theory and combinatorial optimization. Learning Parities under the Uniform Distribution: We study the learnability of parities in the agnostic learning framework of Haussler and Kearns et al. We show that under the uniform distribution, agnostically learning parities reduces to learning parities with random classification noise, commonly referred to as the noisy parity problem. Together with the parity learning algorithm of Blum et al, this gives the first nontrivial algorithm for agnostic learning of parities. We use similar techniques to reduce learning of two other fundamental concept classes under the uniform distribution to learning of noisy parities. Namely, we show that learning of DNF expressions reduces to learning noisy parities of just logarithmic number of variables and learning of k-juntas reduces to learning noisy parities of k variables. Agnostic Learning of Halfspaces: We give an essentially optimal hardness result for agnostic learning of halfspaces over rationals. We show that for any constant ε finding a halfspace that agrees with an unknown function on 1/2+ε fraction of examples is NP-hard even when there exists a halfspace that agrees with the unknown function on 1-ε fraction of examples. This significantly improves on a number of previous hardness results for this problem. We extend the result to ε = 2[superscript-Ω(sqrt{log n})] assuming NP is not contained in DTIME(2[superscript(log n)O(1)]). Majorities of Halfspaces: We show that majorities of halfspaces are hard to PAC-learn using any representation, based on the cryptographic assumption underlying the Ajtai-Dwork cryptosystem. This also implies a hardness result for learning halfspaces with a high rate of adversarial noise even if the learning algorithm can output any efficiently computable hypothesis. Max-Clique, Chromatic Number and Min-3Lin-Deletion: We prove an improved hardness of approximation result for two problems, namely, the problem of finding the size of the largest clique in a graph (also referred to as the Max-Clique problem) and the problem of finding the chromatic number of a graph. We show that for any constant γ > 0, there is no polynomial time algorithm that approximates these problems within factor n/2[superscript(log n)3/4+γ] in an n vertex graph, assuming NP is not contained in BPTIME(2[superscript(log n)O(1)]). This improves the hardness factor of n/2[superscript (log n)1-γ'] for some small (unspecified) constant γ' > 0 shown by Khot. Our main idea is to show an improved hardness result for the Min-3Lin-Deletion problem. An instance of Min-3Lin-Deletion is a system of linear equations modulo 2, where each equation is over three variables. The objective is to find the minimum number of equations that need to be deleted so that the remaining system of equations has a satisfying assignment. We show a hardness factor of 2[superscript sqrt{log n}] for this problem, improving upon the hardness factor of (log n)[superscriptβ] shown by Hastad, for some small (unspecified) constant β > 0. The hardness results for Max-Clique and chromatic number are then obtained using the reduction from Min-3Lin-Deletion as given by Khot. Monotone Multilinear Boolean Circuits for Bipartite Perfect Matching: A monotone Boolean circuit is said to be multilinear if for any AND gate in the circuit, the minimal representation of the two input functions to the gate do not have any variable in common. We show that monotone multilinear Boolean circuits for computing bipartite perfect matching require exponential size. In fact we prove a stronger result by characterizing the structure of the smallest monotone multilinear Boolean circuits for the problem.