Arithmetic Expression Construction

Abstract: When can $n$ given numbers be combined using arithmetic operators from a given subset of ${+, -, \times, \div}$ to obtain a given target number? We study three variations of this problem of Arithmetic Expression Construction: when the expression (1) is unconstrained; (2) has a specified pattern of parentheses and operators (and only the numbers need to be assigned to blanks); or (3) must match a specified ordering of the numbers (but the operators and parenthesization are free). For each of these variants, and many of the subsets of ${+,-,\times,\div}$, we prove the problem NP-complete, sometimes in the weak sense and sometimes in the strong sense. Most of these proofs make use of a "rational function framework" which proves equivalence of these problems for values in rational functions with values in positive integers.

• The paper provides NP-completeness proofs for different variants of arithmetic expression problems using well-defined operator sets and structured expression trees.
• It introduces a rational function framework that reduces multi-variable problems to simpler forms by substituting free variables with sufficiently large integers.
• The work presents pseudopolynomial dynamic programming algorithms for certain NP-hard cases, highlighting the need for heuristic approaches in complex instances.

Arithmetic Expression Construction

Introduction to Arithmetic Expression Construction

The "Arithmetic Expression Construction" paper explores the problem of determining whether a given set of numbers can be combined using arithmetic operators to produce a target number. The paper classifies several variations of the problem according to constraints imposed on expressions. Each variant is analyzed for computational complexity, with proofs of NP-completeness for several operator sets, sometimes indicating weak or strong senses of NP-completeness.

Problem Variants and Complexity

Three main variants of the Arithmetic Expression Construction (AEC) are introduced:

1. Unconstrained Variant: This is the standard variant where no restrictions are imposed on how expressions can be formed using the available operators. For many operator subsets, this variant is shown to be NP-complete.
2. Specified Pattern of Parentheses and Operators (Enforced Operations): In this variant, expressions are required to follow a specific pattern, and only the numbers can be assigned freely to blanks. The paper proves the NP-completeness of this variant for several operator sets.
3. Specified Ordering of Numbers (Enforced Leaves): Here, numbers must follow a specified order in the expression, while operators and parenthesization are free. The variant is similarly proven NP-complete for various operator sets.

Rational Function Framework and Reduction

A significant portion of the paper is devoted to demonstrating a rational function framework. This framework facilitates a polynomial-time equivalence between different versions of the AEC problem across differing numbers of variables. It is pivotal in reducing instances of AEC problems over multiple variables to problems over fewer variables, maintaining problem solvability.

The framework introduces the concept of "sufficiency" for integers within expression instances. This concept enables the removal of free variables by substitution with sufficiently large integers, simplifying the problem to one over constants without losing generality regarding solvability.

Key Theorems and Results

1. Structure Theorem: The paper provides a detailed structure theorem for unconstrained expression trees solving AEC problems. For any solvable instance involving a product of two variables with potentially additional numerical terms, any solution expression tree must have a specific nested structure involving products and sums.
2. NP-Completeness Proofs: Various subsets of arithmetic operators are analyzed:
   • Operator sets such as $\{+, \times\}$, $\{-, \times\}$, and $\{+, \times, \div\}$ are shown to be NP-complete for both the unconstrained and specified pattern variants.
   • The proofs often use reductions from known NP-complete problems like Partition and 3-Partition, leveraging carefully crafted instances of the problem to demonstrate computational hardness.
3. Pseudopolynomial Algorithms: For certain NP-hard subsets, the paper provides pseudopolynomial algorithms, indicating the weak NP-completeness of these problems. These algorithms, based on dynamic programming techniques, further substantiate the complexity classifications.

Implications and Future Directions

The paper’s findings have theoretical implications, affirming several operator combinations as computationally challenging. Practically, these results illuminate the inherent difficulty of building arithmetic expressions under certain constraints, suggesting algorithmic efforts should focus on approximation or heuristic-based approaches for large instances.

Future research directions hinted at in the paper include potential generalizations of the rational function framework to broader computational models and exploration of Arithmetic Minimum Formula Constructions, offering opportunities to further probe combinatorial rigidity in symbolic expression reasoning.