We propose an analytical framework for studying parallel repetition, a basic product operation for one-round twoplayer games. In this framework, we consider a relaxation of the value of projection games. We show that this relaxation is multiplicative with respect to parallel repetition and that it provides a good approximation to the game value. Based on this relaxation, we prove the following improved parallel repetition bound: For every projection game G with value at most ρ, the k-fold parallel repetition G ⊗k has value at most val(G ⊗k )This statement implies a parallel repetition bound for projection games with low value ρ. Previously, it was not known whether parallel repetition decreases the value of such games. This result allows us to show that approximating set cover to within factor (1 − ε) ln n is NP-hard for every ε > 0, strengthening Feige's quasi-NP-hardness and also building on previous work by Moshkovitz and Raz.In this framework, we also show improved bounds for few parallel repetitions of projection games, showing that Raz's counterexample to strong parallel repetition is tight even for a small number of repetitions.Finally, we also give a short proof for the NP-hardness of label cover(1, δ) for all δ > 0, starting from the basic PCP theorem.

The edge expansion of a subset of vertices S ⊆ V in a graph G measures the fraction of edges that leave S. In a d-regular graph, the edge expansion/conductance Φ(S) of a subset S ⊆ V is defined as Φ(S) = |E(S,V \S)| d|S| . Approximating the conductance of small linear sized sets (size δn) is a natural optimization question that is a variant of the well-studied Sparsest Cut problem. However, there are no known algorithms to even distinguish between almost complete edge expansion (Φ(S) = 1 − ε), and close to 0 expansion.In this work, we investigate the connection between Graph Expansion and the Unique Games Conjecture. Specifically, we show the following:-We show that a simple decision version of the problem of approximating small set expansion reduces to Unique Games. Thus if approximating edge expansion of small sets is hard, then Unique Games is hard. Alternatively, a refutation of the UGC will yield better algorithms to approximate edge expansion in graphs.This is the first non-trivial "reverse" reduction from a natural optimization problem to Unique Games.-Under a slightly stronger UGC that assumes mild expansion of small sets, we show that it is UG-hard to approximate small set expansion.-On instances with sufficiently good expansion of small sets, we show that Unique Games is easy by extending the techniques of [4].

Subexponential time approximation algorithms are presented for the U nique G ames and S mall -S et E xpansion problems. Specifically, for some absolute constant c , the following two algorithms are presented. (1) An exp( kn ϵ )-time algorithm that, given as input a k -alphabet unique game on n variables that has an assignment satisfying 1-ϵ c fraction of its constraints, outputs an assignment satisfying 1-ϵ fraction of the constraints. (2) An exp( n ϵ /δ)-time algorithm that, given as input an n -vertex regular graph that has a set S of δ n vertices with edge expansion at most ϵ c , outputs a set S' of at most δ n vertices with edge expansion at most ϵ. subexponential algorithm is also presented with improved approximation to M ax C ut , S parsest C ut , and V ertex C over on some interesting subclasses of instances. These instances are graphs with low threshold rank , an interesting new graph parameter highlighted by this work. Khot's Unique Games Conjecture (UGC) states that it is NP -hard to achieve approximation guarantees such as ours for U nique G ames . While the results here stop short of refuting the UGC, they do suggest that U nique G ames are significantly easier than NP -hard problems such as M ax 3-S at , M ax 3- Lin , L abel C over , and more, which are believed not to have a subexponential algorithm achieving a nontrivial approximation ratio. Of special interest in these algorithms is a new notion of graph decomposition that may have other applications. Namely, it is shown for every ϵ >0 and every regular n -vertex graph G , by changing at most δ fraction of G 's edges, one can break G into disjoint parts so that the stochastic adjacency matrix of the induced graph on each part has at most n ϵ eigenvalues larger than 1-η, where η depends polynomially on ϵ. The subexponential algorithm combines this decomposition with previous algorithms for U nique G ames on graphs with few large eigenvalues [Kolla and Tulsiani 2007; Kolla 2010].

We show a new way to round vector solutions of semidefinite programming (SDP) hierarchies into integral solutions, based on a connection between these hierarchies and the spectrum of the input graph. We demonstrate the utility of our method by providing a new SDP-hierarchy based algorithm for constraint satisfaction problems with 2-variable constraints (2-CSP's).More concretely, we show for every 2-CSP instance ℑ a rounding algorithm for r rounds of the Lasserre SDP hierarchy for ℑ that obtains an integral solution that is at most ε worse than the relaxation's value (normalized to lie in [0, 1]), as long as *