Verifying the Accuracy of Polynomial Approximations in HOL

@inproceedings{Harrison1997VerifyingTA,
  title={Verifying the Accuracy of Polynomial Approximations in HOL},
  author={John Harrison},
  booktitle={TPHOLs},
  year={1997}
}
  • J. Harrison
  • Published in TPHOLs 19 August 1997
  • Computer Science
Many modern algorithms for the transcendental functions rely on a large table of precomputed values together with a low-order polynomial to interpolate between them. In verifying such an algorithm, one is faced with the problem of bounding the error in this polynomial approximation. The most straightforward methods are based on numerical approximations, and are not prima facie reducible to a formal HOL proof. We discuss a technique for proving such results formally in HOL, via the formalization… 

Proving Tight Bounds on Univariate Expressions with Elementary Functions in Coq

TLDR
This paper presents a tactic for the Coq proof assistant that is designed to automatically and formally prove bounds on univariate expressions based on a formalization of floating-point and interval arithmetic, associated with an on-the-fly computation of Taylor expansions.

Proving Tight Bounds on Univariate Expressions in Coq

TLDR
A tactic for the Coq proof assistant that is designed to automatically and formally prove bounds on univariate expressions is presented, based on a kernel of floating-point and interval arithmetic, associated with an on-the-fly computation of Taylor expansions.

A Decision Procedure for Univariate Real Polynomials in Isabelle/HOL

TLDR
An Isabelle/HOL proof method was implemented to prove interesting statements about the number of real roots of a univariate real polynomial and related properties such as non-negativity and monotonicity.

Mechanical Verification of a Square Root Algorithm Using Taylor's Theorem

TLDR
The IBM Power4 processor uses Chebyshev polynomials to calculate square root and the correctness of this algorithm is verified using the ACL2(r) theorem prover, which proves Taylor's theorem by way of non-standard analysis.

Floating Point Verification in HOL Light: The Exponential Function

  • J. Harrison
  • Computer Science
    Formal Methods Syst. Des.
  • 1997
TLDR
A machine-checked verification of an algorithm for computing the exponential function in IEEE-754 standard binary floating point arithmetic, developed logically from first principles using the HOL Light prover, which guarantees strict adherence to simple rules of inference while allowing the user to perform proofs using higher-level derived rules.

A Formally-Verified Decision Procedure for Univariate Polynomial Computation Based on Sturm's Theorem

Sturm's Theorem is a well-known result in real algebraic geometry that provides a function that computes the number of roots of a univariate polynomial in a semiopen interval. This paper presents a

Formal Verification of Floating Point Trigonometric Functions

TLDR
This paper describes in some depth the formal verification of the sin and cos functions, including the initial range reduction step, covering both pure mathematics and the detailed analysis of floating point rounding.

Complex quantifier elimination in HOL

TLDR
This work implements, as a HOL derived inference rule, a decision method for the first order algebraic theory of C based on quantifier elimination and implements a more efficient semidecision procedure for the universal fragment based on Gr öbner bases.

Provably faithful evaluation of polynomials

TLDR
A formalization of floating-point numbers and rounding modes in the Program Verification System (PVS) based on a well-known formalization in the proof assistant Coq to provide sufficient conditions that formally guarantee that the floating- point computation of a polynomial evaluation is faithful.

Formally-Verified Decision Procedures for Univariate Polynomial Computation Based on Sturm’s and Tarski’s Theorems

TLDR
A formalization of two decision procedures, based on Sturm’s and Tarski theorems, that determines satisfiability of a single polynomial relation over an interval and the soundness and completeness properties of these decision procedures are formally verified in PVS are presented.

References

SHOWING 1-10 OF 13 REFERENCES

Floating Point Verification in HOL Light: The Exponential Function

  • J. Harrison
  • Computer Science
    Formal Methods Syst. Des.
  • 1997
TLDR
A machine-checked verification of an algorithm for computing the exponential function in IEEE-754 standard binary floating point arithmetic, developed logically from first principles using the HOL Light prover, which guarantees strict adherence to simple rules of inference while allowing the user to perform proofs using higher-level derived rules.

Theorem Proving with the Real Numbers

TLDR
HOL extends Church's system by allowing also 'type variables' which give a form of polymorphism, and is capable not only of deciding whether it has a type, but inferring a most general type for it if it does.

Computability in analysis and physics

TLDR
This book represents the first treatment of computable analysis at the graduate level within the tradition of classical mathematical reasoning and is sufficiently detailed to provide an introduction to research in this area.

Table-driven implementation of the exponential function in IEEE floating-point arithmetic

TLDR
Algorithms and implementation details for the exponential function in both single- and double-precision of IEEE 754 arithmetic are presented here and are provably accurate to within 0.54 ulp as long as the final result does not underflow.

Computational Logic

TLDR
This paper presents a meta-modelling simulation of the response of the immune system to changes in the environment through the course of natural selection.

Sur le calcul effectif des polynomes d’approximation de Tchebichef

  • Comptes Rendus Hebdomadaires des Séances de l’Académie des Sciences,
  • 1934

Polynomials of best uniform approximation to certain rational functions

<= (&23 / &27) * inv(&2 pow 33) References

  • <= (&23 / &27) * inv(&2 pow 33) References