Strongly Polynomial Bounds in Linear Programming and Smale's 9th Problem

This preprint provides an exhaustive 9-page mathematical monograph on Smale's 9th Problem (Steve Smale, 2000), dedicated to the existence of a strongly polynomial-time algorithm for Linear Programming (LP). While the ellipsoid method (Khachiyan, 1979) and projective interior-point methods (Karmarkar, 1984) established that linear programming is solvable in weakly polynomial time (where complexity depends on the binary bit-length L), finding an algorithm whose arithmetic operations depend solely on the matrix dimensions m and n in the real RAM model remains one of the greatest open problems in optimization and theoretical computer science: ∃? Algorithme avec T(A, b, c) ≤ O((m + n)C) opérations arithmétiques dans ℝ This treatise develops a comprehensive, non-elliptical exposition of primal-dual complementary slackness, Éva Tardos' landmark theorem (1986) for matrices with bounded subdeterminants Δ, the Vavasis-Ye Layered Least- Squares (LLS) interior-point algorithm (1996) parameterized by χ̄(A), and tropical geometry curvature barriers (Allamigeon et al., 2018). Key Mathematical Results & Contributions Primal-Dual Duality Architecture: Exact algebraic derivation of the duality gap identity ⟨c, x⟩ − ⟨b, y⟩ = ⟨c − ATy, x⟩, establishing weak duality and complementary slackness equivalence. Éva Tardos' Landmark Theorem (1986): Step-by-step proof that LP is strongly polynomial with runtime O(poly(m, n, log Δ)) when subdeterminants satisfy Δ = maxB | det(AB)|, completely independent of the bit-lengths of b and c. Vavasis-Ye Layered Least-Squares (LLS) Method (1996): Detailed analysis of continuous path-following in ill-conditioned regimes, proving termination in O(n3.5 log(χ̄(A) + n)) iterations. Tropical Geometry Curvature Barriers: Structural study of log-barrier central path geometries showing exponential bendings for generic interior-point paths. 100% Machine-Checked Verification in Lean 4: Duality gap identity, weak duality theorem, complementary slackness equivalence, unimodular basis integer preservation, and logarithmic barrier contraction are machine-certified with 0 axioms, 0 linter warnings, and 0 sorry placeholders via the Lean 4 interactive theorem prover and Mathlib. Repository and Verification Artifacts The companion machine-checked code and formal verification artifacts are publicly hosted on GitHub:https://github.com/flouzzy/smale-problemsInteractive Portal: https://maths-proofs.edounze.com/#smale-09-linear-programming Primary MSC (2020): 90C05, 68Q25, 90C51, 68V20, 15A15, 52B12.Keywords: Smale's 9th Problem, Linear Programming, Strongly Polynomial Complexity, Weakly Polynomial, Tardos' Theorem, Interior-Point Methods, Central Path, Vavasis-Ye Algorithm, Duality Gap, Formal Verification, Lean 4, Mathlib.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-08-28
DOI
https://doi.org/10.5281/zenodo.22148791
Primary Topic
Advanced Optimization Algorithms Research
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

Strongly Polynomial Bounds in Linear Programming and Smale's 9th Problem

Charles EDOU NZE
Zenodo (CERN European Organization for Nuclear Research)
Advanced Optimization Algorithms Research
preprint

Strongly Polynomial Bounds in Linear Programming and Smale's 9th Problem

Charles EDOU NZE
preprint en

Abstract

This preprint provides an exhaustive 9-page mathematical monograph on Smale's 9th Problem (Steve Smale, 2000), dedicated to the existence of a strongly polynomial-time algorithm for Linear Programming (LP). While the ellipsoid method (Khachiyan, 1979) and projective interior-point methods (Karmarkar, 1984) established that linear programming is solvable in weakly polynomial time (where complexity depends on the binary bit-length L), finding an algorithm whose arithmetic operations depend solely on the matrix dimensions m and n in the real RAM model remains one of the greatest open problems in optimization and theoretical computer science: ∃? Algorithme avec T(A, b, c) ≤ O((m + n)C) opérations arithmétiques dans ℝ This treatise develops a comprehensive, non-elliptical exposition of primal-dual complementary slackness, Éva Tardos' landmark theorem (1986) for matrices with bounded subdeterminants Δ, the Vavasis-Ye Layered Least- Squares (LLS) interior-point algorithm (1996) parameterized by χ̄(A), and tropical geometry curvature barriers (Allamigeon et al., 2018). Key Mathematical Results & Contributions Primal-Dual Duality Architecture: Exact algebraic derivation of the duality gap identity ⟨c, x⟩ − ⟨b, y⟩ = ⟨c − ATy, x⟩, establishing weak duality and complementary slackness equivalence. Éva Tardos' Landmark Theorem (1986): Step-by-step proof that LP is strongly polynomial with runtime O(poly(m, n, log Δ)) when subdeterminants satisfy Δ = maxB | det(AB)|, completely independent of the bit-lengths of b and c. Vavasis-Ye Layered Least-Squares (LLS) Method (1996): Detailed analysis of continuous path-following in ill-conditioned regimes, proving termination in O(n3.5 log(χ̄(A) + n)) iterations. Tropical Geometry Curvature Barriers: Structural study of log-barrier central path geometries showing exponential bendings for generic interior-point paths. 100% Machine-Checked Verification in Lean 4: Duality gap identity, weak duality theorem, complementary slackness equivalence, unimodular basis integer preservation, and logarithmic barrier contraction are machine-certified with 0 axioms, 0 linter warnings, and 0 sorry placeholders via the Lean 4 interactive theorem prover and Mathlib. Repository and Verification Artifacts The companion machine-checked code and formal verification artifacts are publicly hosted on GitHub:https://github.com/flouzzy/smale-problemsInteractive Portal: https://maths-proofs.edounze.com/#smale-09-linear-programming Primary MSC (2020): 90C05, 68Q25, 90C51, 68V20, 15A15, 52B12.Keywords: Smale's 9th Problem, Linear Programming, Strongly Polynomial Complexity, Weakly Polynomial, Tardos' Theorem, Interior-Point Methods, Central Path, Vavasis-Ye Algorithm, Duality Gap, Formal Verification, Lean 4, Mathlib.

Zenodo (CERN European Organization for Nuclear Research)
Advanced Optimization Algorithms Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.