The author explores the computation of approximate roots for monotone functions, highlighting conditions and methods for efficient root-finding.
Resumo
The content delves into the computation of approximate roots for monotone functions. It discusses conditions under which polynomiality results extend to multi-dimensional functions and presents algorithms for efficient root-finding. The analysis covers various scenarios, including diagonal and ex-diagonal monotonicity conditions, providing insights into the complexity of finding roots in higher dimensions.
Computing approximate roots of monotone functions
Estatísticas
Given a function f : [a, b] → R, if f(a) < 0 and f(b) > 0 and f is continuous, the Intermediate Value Theorem implies that f has a root in [a, b].
The number of required evaluations using the bisection method is polynomial in the number of accuracy digits.
For a two-dimensional function with a single monotonicity condition, the number of required evaluations is polynomial in accuracy.
For d-dimensional functions satisfying all d2 − d "ex-diagonal" monotonicity conditions, the number of evaluations is polynomial in accuracy.
Citações
"In general, finding an approximate root might require an exponential number of evaluations even for a two-dimensional function."
"If f satisfies only d2 − d − 2 ex-diagonal conditions, then the number of required evaluations may be exponential in accuracy."
Is there an algorithm that requires a number of evaluations that is polynomial in the number of digits
In the context provided, the algorithm presented in Theorem 1.3 demonstrates that for a two-dimensional positive-switching function with certain monotonicity conditions, an approximate root can be found using O(log2(1/ε)) evaluations. This result shows that there is indeed an algorithm that requires a number of evaluations polynomial in the number of digits when specific conditions are met.
Does every Lipschitz-continuous positive-switching function need Ω((L/ε)d−1) evaluations
Not every Lipschitz-continuous positive-switching function needs Ω((L/ε)d−1) evaluations. The lower bound mentioned in Proposition 1.5 applies to functions satisfying d2 − 2 monotonicity conditions, including all diagonal and ex-diagonal conditions except for two ex-diagonal ones. For such functions, the number of required evaluations may be exponential in (L/ε).
How does Tarski's fixed-point theorem relate to finding approximate roots
Tarski's fixed-point theorem is closely related to finding approximate roots through its connection with Brouwer's fixed-point theorem and Miranda's root theorem. Tarski's theorem states that every order-preserving function from a complete lattice to itself has a fixed point. By discretizing the value space into a lattice and applying Tarski's theorem conceptually, Fearnley et al., Chen and Li have shown algorithms for finding fixed points efficiently based on this principle.
0
Visualizar esta Página
Gerar com IA indetectável
Traduzir para Outro Idioma
Pesquisa Acadêmica
Sumário
Analyzing Approximate Roots of Monotone Functions
Computing approximate roots of monotone functions
Is there an algorithm that requires a number of evaluations that is polynomial in the number of digits
Does every Lipschitz-continuous positive-switching function need Ω((L/ε)d−1) evaluations
How does Tarski's fixed-point theorem relate to finding approximate roots