arXiv subjects

Rémi Imbach

Publications and source records attributed to Rémi Imbach.

11 recordsLinked to original sources

arXiv2026

Data-efficient extraction of optical properties from 3D Monte Carlo TPSFs using Bi-LSTM transfer learning

Time-Resolved Spectroscopy (TRS) is a powerful modality for non-invasive characterization of turbid media. However, extracting optical properties, absorption $μ_a$ and reduced scattering $μ_s'$, from 3D stochastic measurements remains computationally expensive for real-time applications. In this paper, we propose a data-efficient, physics-informed transfer learning strategy using a Bidirectional Long Short-Term Memory (Bi-LSTM) network. By leveraging a fast deterministic solver to establish a physical prior before fine-tuning on a restricted set of 3D Monte Carlo simulations, our model successfully bridges the analytical-to-stochastic domain gap. The proposed method eliminates the systematic bias of analytical models while maintaining a competitive error with near-instantaneous inference time.

math.NA↗

arXiv2023

Fast evaluation and root finding for polynomials with floating-point coefficients

Evaluating or finding the roots of a polynomial $f(z) = f_0 + \cdots + f_d z^d$ with floating-point number coefficients is a ubiquitous problem. By using a piecewise approximation of $f$ obtained with a careful use of the Newton polygon of $f$, we improve state-of-the-art upper bounds on the number of operations to evaluate and find the roots of a polynomial. In particular, if the coefficients of $f$ are given with $m$ significant bits, we provide for the first time an algorithm that finds all the roots of $f$ with a relative condition number lower than $2^m$, using a number of bit operations quasi-linear in the bit-size of the floating-point representation of $f$. Notably, our new approach handles efficiently polynomials with coefficients ranging from $2^{-d}$ to $2^d$, both in theory and in practice.

cs.SC↗

arXiv2022

Accelerated Subdivision for Clustering Roots of Polynomials given by Evaluation Oracles

In our quest for the design, the analysis and the implementation of a subdivision algorithm for finding the complex roots of univariate polynomials given by oracles for their evaluation, we present sub-algorithms allowing substantial acceleration of subdivision for complex roots clustering for such polynomials. We rely on Cauchy sums which approximate power sums of the roots in a fixed complex disc and can be computed in a small number of evaluations --polylogarithmic in the degree. We describe root exclusion, root counting, root radius approximation and a procedure for contracting a disc towards the cluster of root it contains, called $\varepsilon$-compression. To demonstrate the efficiency of our algorithms, we combine them in a prototype root clustering algorithm. For computing clusters of roots of polynomials that can be evaluated fast, our implementation competes advantageously with user's choice for root finding, MPsolve.

cs.SC↗

arXiv2021

Root Radii and Subdivision for Polynomial Root-Finding

We depart from our approximation of 2000 of all root radii of a polynomial, which has readily extended Sch{ö}nhage's efficient algorithm of 1982 for a single root radius. We revisit this extension, advance it, based on our simple but novel idea, and yield significant practical acceleration of the known near optimal subdivision algorithms for complex and real root-finding of user's choice. We achieve this by means of significant saving of exclusion tests and Taylor's shifts, which are the bottleneck of subdivision root-finders. This saving relies on our novel recipes for the initialization of root-finding iterations of independent interest. We demonstrate our practical progress with numerical tests, provide extensive analysis of the resulting algorithms, and show that, like the preceding subdivision root-finders, they support near optimal Boolean complexity bounds.

cs.SC↗

arXiv2019

Polynomial root clustering and explicit deflation

We seek complex roots of a univariate polynomial $P$ with real or complex coefficients. We address this problem based on recent algorithms that use subdivision and have a nearly optimal complexity. They are particularly efficient when only roots in a given Region Of Interest (ROI) are sought. We propose two improvements for root finders. The first one is applied to polynomials having only real coefficients; their roots are either real or appear in complex conjugate pairs. We show how to adapt the subdivision scheme to focus the computational effort on the imaginary positive part of the ROI. In our second improvement we deflate $P$ to decrease its degree and the arithmetic cost of the subdivision.

cs.SC↗

arXiv2019

New practical advances in polynomial root clustering

We report an ongoing work on clustering algorithms for complex roots of a univariate polynomial $p$ of degree $d$ with real or complex coefficients. As in their previous best subdivision algorithms our root-finders are robust even for multiple roots of a polynomial given by a black box for the approximation of its coefficients, and their complexity decreases at least proportionally to the number of roots in a region of interest (ROI) on the complex plane, such as a disc or a square, but we greatly strengthen the main ingredient of the previous algorithms. Namely our new counting test essentially amounts to the evaluation of a polynomial $p$ and its derivative $p'$, which is a major benefit, e.g., for sparse polynomials $p$. Moreover with evaluation at about $\log(d)$ points (versus the previous record of order $d$) we output correct number of roots in a disc whose contour has no roots of $p$ nearby. Moreover we greatly soften the latter requirement versus the known subdivision algorithms. Our second and less significant contribution concerns subdivision algorithms for polynomials with real coefficients. Our tests demonstrate the power of the proposed algorithms.

cs.SC↗

arXiv2019

Clustering Complex Zeros of Triangular Systems of Polynomials

This paper gives the first algorithm for finding a set of natural $ε$-clusters of complex zeros of a triangular system of polynomials within a given polybox in $\mathbb{C}^n$, for any given $ε>0$. Our algorithm is based on a recent near-optimal algorithm of Becker et al (2016) for clustering the complex roots of a univariate polynomial where the coefficients are represented by number oracles. Our algorithm is numeric, certified and based on subdivision. We implemented it and compared it with two well-known homotopy solvers on various triangular systems. Our solver always gives correct answers, is often faster than the homotopy solver that often gives correct answers, and sometimes faster than the one that gives sometimes correct results.

cs.CG↗

arXiv2018

Implementation of a Near-Optimal Complex Root Clustering Algorithm

We describe Ccluster, a software for computing natural $ε$-clusters of complex roots in a given box of the complex plane. This algorithm from Becker et al.~(2016) is near-optimal when applied to the benchmark problem of isolating all complex roots of an integer polynomial. It is one of the first implementations of a near-optimal algorithm for complex roots. We describe some low level techniques for speeding up the algorithm. Its performance is compared with the well-known MPSolve library and Maple.

cs.MS↗

arXiv2016

A Subdivision Solver for Systems of Large Dense Polynomials

We describe here the package {\tt subdivision\\_solver} for the mathematical software {\tt SageMath}. It provides a solver on real numbers for square systems of large dense polynomials. By large polynomials we mean multivariate polynomials with large degrees, which coefficients have large bit-size. While staying robust, symbolic approaches to solve systems of polynomials see their performances dramatically affected by high degree and bit-size of input polynomials.Available numeric approaches suffer from the cost of the evaluation of large polynomials and their derivatives.Our solver is based on interval analysis and bisections of an initial compact domain of $\R^n$ where solutions are sought. Evaluations on intervals with Horner scheme is performed by the package {\tt fast\\_polynomial} for {\tt SageMath}.The non-existence of a solution within a box is certified by an evaluation scheme that uses a Taylor expansion at order 2, and existence and uniqueness of a solution within a box is certified with krawczyk operator.The precision of the working arithmetic is adapted on the fly during the subdivision process and we present a new heuristic criterion to decide if the arithmetic precision has to be increased.

cs.MS↗

arXiv2016

A Robust and Efficient Method for Solving Point Distance Problems by Homotopy

The goal of Point Distance Solving Problems is to find 2D or 3D placements of points knowing distances between some pairs of points. The common guideline is to solve them by a numerical iterative method (\emph{e.g.} Newton-Raphson method). A sole solution is obtained whereas many exist. However the number of solutions can be exponential and methods should provide solutions close to a sketch drawn by the user.Geometric reasoning can help to simplify the underlying system of equations by changing a few equations and triangularizing it.This triangularization is a geometric construction of solutions, called construction plan. We aim at finding several solutions close to the sketch on a one-dimensional path defined by a global parameter-homotopy using a construction plan. Some numerical instabilities may be encountered due to specific geometric configurations. We address this problem by changing on-the-fly the construction plan.Numerical results show that this hybrid method is efficient and robust.

cs.CG↗

arXiv2015

Numeric certified algorithm for the topology of resultant and discriminant curves

Let $\mathcal C$ be a real plane algebraic curve defined by the resultant of two polynomials (resp. by the discriminant of a polynomial). Geometrically such a curve is the projection of the intersection of the surfaces $P(x,y,z)=Q(x,y,z)=0$ (resp. $P(x,y,z)=\frac{\partial P}{\partial z}(x,y,z)=0$), and generically its singularities are nodes (resp. nodes and ordinary cusps). State-of-the-art numerical algorithms compute the topology of smooth curves but usually fail to certify the topology of singular ones. The main challenge is to find practical numerical criteria that guarantee the existence and the uniqueness of a singularity inside a given box $B$, while ensuring that $B$ does not contain any closed loop of $\mathcal{C}$. We solve this problem by first providing a square deflation system, based on subresultants, that can be used to certify numerically whether $B$ contains a unique singularity $p$ or not. Then we introduce a numeric adaptive separation criterion based on interval arithmetic to ensure that the topology of $\mathcal C$ in $B$ is homeomorphic to the local topology at $p$. Our algorithms are implemented and experiments show their efficiency compared to state-of-the-art symbolic or homotopic methods.

cs.CG↗