An efficient algorithm for the complex roots problem
Improving their own algorithm of 1994, the authors present an algorithm to isolate all complex roots of an \(n\)-degree polynomial \(f=z^n+\sum_{i=0}^{n-1} c_i z^i\) with specified precision \(2^{-\mu}\) in \(O(n \log^5 n \log B)\) arithmetic operations, where \(B= \chi + \mu\), \(\chi=\max \{ m, n\}\), \(|c_i|< 2^m, i=0,\ldots, n-1\). The algorithm improves the upper bound for the complex roots problem due to \textit{V. Y. Pan} [Comput. Math. Appl. 14, 285-304 (1987; Zbl 0632.65052)] and approximates its complexity to those of basic operations such as interpolation and evaluation. The algorithm hinges on a balance splitting technique in which an \(n\)-degree polynomial is translated to a coordinate system where it can be approximately factored into smaller degree polynomials (of degree \(\leq n/2\)). Such a technique is based on the notion of splitting set of points for complex polynomials developed by the authors, which may be seen as a generalization of the splitting point idea, used to obtain a geometrically balanced set of roots for totally real polynomials. A remarkable property of the algorithm presented is that it does not need any assumption about the root separation of \(f\). The authors also show that their splitting procedure works without modification in the Boolean model of arithmetic. The complexity in the Boolean model is only increased by a multiplicative factor (explicitly determined). The paper also discusses a parallel implementation of the algorithm. It is not clear whether the algorithm can be translated to an efficient numerical routine, but its relatively simple description seems to indicate that it is possible to simplify the presentation, which may lead to an actual implementation.
- A refined model of computation for continuous problems
- Partial fraction decomposition in \(\mathbb{C}(z)\) and simultaneous Newton iteration for factorization in \(\mathbb{C}^{[z]}\)
- Why does information-based complexity use the real number model?
- Some speed-ups and speed limits for real algebraic geometry
- A near-optimal subdivision algorithm for complex root isolation based on the Pellet test and Newton iteration
- An adaptive subdivision method for root finding of univariate polynomials
- On solving univariate sparse polynomials in logarithmic time
- Graeffe's, Chebyshev-like, and Cardinal's processes for splitting a polynomial into factors
- Algorithms for quaternion polynomial root-finding
- Algorithms for \(n\)-th root approximation.
- Near optimal subdivision algorithms for real root isolation
- Univariate polynomials, nearly optimal algorithms for factorization and rootfinding
- Advice coins for classical and quantum computation
- On the parallel arithmetic complexity of the root-finding problem
- A Fast Parallel Algorithm for Determining All Roots of a Polynomial with Real Roots
- scientific article; zbMATH DE number 1263360 (Why is no real title available?)
- Tight space-noise tradeoffs in computing the ergodic measure
- An algorithm for computing the roots of a complex polynomial
- scientific article; zbMATH DE number 880382 (Why is no real title available?)
- Newton's method and the computational complexity of the fundamental theorem of algebra
- Spectral norm of a symmetric tensor and its computation
- Inapproximability of the independent set polynomial in the complex plane
- An efficient algorithm for computing the roots of general quadratic, cubic and quartic equations
- Computational arithmetic geometry. I: Sentences nearly in the polynomial hierarchy
- On the geometry of Graeffe iteration
- Validated Root Enclosures for Interval Polynomials with Multiplicities
- CCOMP: an efficient algorithm for complex roots computation of determinantal equations
- Globally convergent, iterative path-following for algebraic equations
- A fast and stable algorithm for splitting polynomials
- Root finding with threshold circuits
- Univariate polynomials: Nearly optimal algorithms for numerical factorization and root-finding
- Low-temperature sampling on sparse random graphs
- A parallel algorithm for finding roots of a complex polynomial
This page was built for publication: An efficient algorithm for the complex roots problem
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2565192)