We consider the Min-r-Lin(ℤ_m) problem: given a system S of length-r linear equations modulo m, find Z ⊆ S of minimum cardinality such that S-Z is satisfiable. The problem is NP-hard and UGC-hard to approximate in polynomial time within any constant factor even when r = m = 2. We focus on parameterized approximation with solution size as the parameter. Dabrowski, Jonsson, Ordyniak, Osipov and Wahlström [SODA-2023] showed that Min-r-Lin(ℤ_m) is in FPT if m is prime (i.e. ℤ_m is a field), and it is W[1]-hard if m is not a prime power. We show that Min-r-Lin(ℤ_{pⁿ}) is FPT-approximable within a factor of 2 for every prime p and integer n ≥ 2. This implies that Min-2-Lin(ℤ_m), m ∈ ℤ^+, is FPT-approximable within a factor of 2ω(m) where ω(m) counts the number of distinct prime divisors of m. The high-level idea behind the algorithm is to solve tighter and tighter relaxations of the problem, decreasing the set of possible values for the variables at each step. When working over ℤ_{pⁿ} and viewing the values in base-p, one can roughly think of a relaxation as fixing the number of trailing zeros and the least significant nonzero digits of the values assigned to the variables. To solve the relaxed problem, we construct a certain graph where solutions can be identified with a particular collection of cuts. The relaxation may hide obstructions that will only become visible in the next iteration of the algorithm, which makes it difficult to find optimal solutions. To deal with this, we use a strategy based on shadow removal [Marx & Razgon, STOC-2011] to compute solutions that (1) cost at most twice as much as the optimum and (2) allow us to reduce the set of values for all variables simultaneously. We complement the algorithmic result with two lower bounds, ruling out constant-factor FPT-approximation for Min-3-Lin(R) over any nontrivial ring R and for Min-2-Lin(R) over some finite commutative rings R.

In this thesis we study the computational complexity of MinCSP - an optimization version of the Constraint Satisfaction Problem (CSP). The input to a MinCSP is a set of variables and constraints applied to these variables, and the goal is to assign values (from a fixed domain) to the variables while minimizing the solution cost, i.e. the number of unsatisfied constraints. We are specifically interested in MinCSP with infinite domains of values. Infinite-domain MinCSPs model fundamental optimization problems in computer science and are of particular relevance to artificial intelligence, especially temporal and spatial reasoning. The usual way to study computational complexity of CSPs is to restrict the types of constraints that can be used in the inputs, and either construct fast algorithms or prove lower bounds on the complexity of the resulting problems.

The vast majority of interesting MinCSPs are NP-hard, so standard complexity-theoretic assumptions imply that we cannot find exact solutions to all inputs of these problems in polynomial time with respect to the input size. Hence, we need to relax at least one of the three requirements above, opting for either approximate solutions, solving some inputs, or using super-polynomial time. Parameterized algorithms exploits the latter two relaxations by identifying some common structure of the interesting inputs described by some parameter, and then allowing super-polynomial running times with respect to that parameter. Such algorithms are feasible for inputs of any size whenever the parameter value is small. For MinCSP, a natural parameter is optimal solution cost. We also study parameterized approximation algorithms, where the requirement for exact solutions is also relaxed.

We present complete complexity classifications for several important classes of infinite-domain constraints. These are simple temporal constraints and interval constraints, which have notable applications in temporal reasoning in AI, linear equations over finite and infinite fields as well as some commutative rings (e.g., the rationals and the integers), which are of fundamental theoretical importance, and equality constraints, which are closely related to connectivity problems in undirected graphs and form the basis of studying first-order definable constraints over infinite domains. In all cases, we prove results as follows: we fix a (possibly infinite) set of allowed constraint types C, and for every finite subset of C, determine whether MinCSP(), i.e., MinCSP restricted to the constraint types in , is fixed-parameter tractable, i.e. solvable in f(k) · poly(n) time, where k is the parameter, n is the input size, and f is any function that depends solely on k. To rule out such algorithms, we prove lower bounds under standard assumptions of parameterized complexity. In all cases except simple temporal constraints, we also provide complete classifications for fixed-parameter time constant-factor approximation.

Checking whether a system of linear equations is consistent is a basic computational problem with ubiquitous applications. When dealing with inconsistent systems, one may seek an assignment that minimizes the number of unsatisfied equations. This problem is NP-hard and UGC-hard to approximate within any constant even for two-variable equations over the two-element field. We study this problem from the point of view of parameterized complexity, with the parameter being the number of unsatisfied equations. We consider equations defined over Euclidean domains—a family of commutative rings that generalize finite and infinite fields including the rationals, the ring of integers and many other structures. We show that if every equation contains at most two variables, the problem is fixed-parameter tractable. This generalizes many eminent graph separation problems such as Bipartization, Multiway Cut and Multicut parameterized by the size of the cutset. To complement this, we show that the problem is W[1]-hard when three or more variables are allowed in an equation, as well as for many commutative rings that are not Euclidean domains. On the technical side, we introduce the notion of important balanced subgraphs, generalizing important separators of Marx [Theor. Comput. Sci. 2006] to the setting of biased graphs. Furthermore, we use recent results on parameterized MinCSP [Kim et al., SODA 2021] to efficiently solve a generalization of Multicut with disjunctive cut requests.

Constraint satisfaction problems form a nicely behaved class of problems that lends itself to complexity classification results. From the point of view of parameterized complexity, a natural task is to classify the parameterized complexity of MinCSP problems parameterized by the number of unsatisfied constraints. In other words, we ask whether we can delete at most k constraints, where k is the parameter, to get a satisfiable instance. In this work, we take a step towards classifying the parameterized complexity for an important infinite-domain CSP: Allen’s interval algebra (IA). This CSP has closed intervals with rational endpoints as domain values and employs a set A of 13 basic comparison relations such as "precedes" or "during" for relating intervals. IA is a highly influential and well-studied formalism within AI and qualitative reasoning that has numerous applications in, for instance, planning, natural language processing and molecular biology. We provide an FPT vs. W[1]-hard dichotomy for MinCSP(Γ) for all Γ ⊆ A. IA is sometimes extended with unions of the relations in A or first-order definable relations over A, but extending our results to these cases would require first solving the parameterized complexity of Directed Symmetric Multicut, which is a notorious open problem. Already in this limited setting, we uncover connections to new variants of graph cut and separation problems. This includes hardness proofs for simultaneous cuts or feedback arc set problems in directed graphs, as well as new tractable cases with algorithms based on the recently introduced flow augmentation technique. Given the intractability of MinCSP(A) in general, we then consider (parameterized) approximation algorithms. We first show that MinCSP(A) cannot be polynomial-time approximated within any constant factor and continue by presenting a factor-2 fpt-approximation algorithm. Once again, this algorithm has its roots in flow augmentation. 

We study the parameterized complexity of MinCSP for so-called equality languages, i.e., for finite languages over an infinite domain such as ℕ, where the relations are defined via first-order formulas whose only predicate is =. This is an important class of languages that forms the starting point of all study of infinite-domain CSPs under the commonly used approach pioneered by Bodirsky, i.e., languages defined as reducts of finitely bounded homogeneous structures. Moreover, MinCSP over equality languages forms a natural class of optimisation problems in its own right, covering such problems as Edge Multicut, Steiner Multicut and (under singleton expansion) Edge Multiway Cut. We classify MinCSP(Γ) for every finite equality language Γ, under the natural parameter, as either FPT, W[1]-hard but admitting a constant-factor FPT-approximation, or not admitting a constant-factor FPT-approximation unless FPT=W[2]. In particular, we describe an FPT case that slightly generalises Multicut, and show a constant-factor FPT-approximation for Disjunctive Multicut, the generalisation of Multicut where the "cut requests" come as disjunctions over O(1) individual cut requests s_i ≠ t_i. We also consider singleton expansions of equality languages, enriching an equality language with the capability for assignment constraints (x = i) for either a finite or infinitely many constants i, and fully characterize the complexity of the resulting MinCSP.

Many computational problems arising in, for instance, artificial intelligence can be realized as infinite-domain constraint satisfaction problems (CSPs) based on partition schemes: a set of pairwise disjoint binary relations (containing the equality relation) whose union spans the underlying domain and which is closed under converse. We first consider partition schemes that contain an acyclic order and where the constraint language contains all unions of the basic relations; such CSPs are frequently occurring in e.g. temporal and spatial reasoning. We identify properties of such orders which, when combined, are sufficient to establish NP-hardness of the CSP and strong lower bounds under the exponential-time hypothesis, even for degree-bounded problems. This result explains, in a uniform way, many existing hardness results from the literature, and shows that it is impossible to obtain subexponential time algorithms unless the exponential-time hypothesis fails. However, some of these problems (including several important temporal problems), despite likely not being solvable in subexponential time, admit non-trivial improved exponential-time algorithm, and we present a novel improved algorithm for RCC-8 and related formalisms.