EUROPEAN JOURNAL OF PURE AND APPLIED MATHEMATICS 2025, Vol. 18, Issue 3, Article Number 6440 ISSN 1307-5543 – ejpam.com Published by New York Business Global Fixed Point Theory in MR-Metric Spaces: Fundamental Theorems and Applications to Integral Equations and Neutron Transport Tariq A. Qawasmeh1, Abed Al-Rahman M. Malkawi1,∗ 1 Department of Mathematics, Faculty of Arts and Science, Amman Arab University, Amman 11953, Jordan Abstract. This paper establishes a comprehensive framework for fixed point theory in MR-metric spaces, a generalization of standard metric spaces that incorporates three-point relations. We present four fundamental theorems: (i) A Banach contraction principle with optimal contraction constant k < 1 3R (ii) A solvability theorem for Fredholm-type integral equations (iii) A Krasnoselskii-type hybrid fixed point theorem (iv) A Leray-Schauder alternative for generalized contractions The theoretical results are applied to: • Nonlinear integral equations in neutron transport theory • Optimization problems in neural networks • Boundary value problems for nonlinear ODEs Key innovations include the development of error estimates in the MR-metric framework and the derivation of precise existence conditions for operator equations. The work bridges theoretical mathematics with practical applications in physics and machine learning. 2020 Mathematics Subject Classifications: 47H10, 54E50, 45G10, 34B15 Key Words and Phrases: MR−metric MR-metric spaces, fixed point theory, Banach contrac- tion principle, integral equations, neutron transport theory, neural network optimization ∗Corresponding author. DOI: https://doi.org/10.29020/nybg.ejpam.v18i3.6440 Email addresses: t.qawasmeh@aau.edu.jo (T. Qawasmeh), a.malkawi@aau.edu.jo, math.malkawi@gmail.com (A. Malkawi) https://www.ejpam.com 1 Copyright: © 2025 The Author(s). (CC BY-NC 4.0) T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 2 of 20 1. Introduction The study of fixed point theory in generalized metric spaces has been a vibrant area of research since Banach’s seminal contraction mapping principle [1]. Recent developments have extended this theory to various abstract spaces, including partial metric spaces [2], b-metric spaces [3], and modular metric spaces [4]. 1.1. MR-Metric Spaces The MR-metric space (X,M), first introduced in [5], provides a framework where the distance function M : X3 → [0,∞) simultaneously measures three-point relations. This structure proves particularly valuable when analyzing: • Systems with ternary interactions • Problems where pairwise distances are insufficient • Operator equations with multi-point constraints 1.2. Contributions Our main contributions are: (i) Theoretical Foundations: • Complete proofs of four fundamental fixed point theorems • Optimal contraction constants in the MR-metric setting • Error estimates for iterative methods (ii) Applications: • New existence results for neutron transport equations • Convergence conditions for neural network training • Solvability criteria for Hammerstein integral equations (iii) Computational Implications: • Layer-wise learning rate bounds in deep learning • Iterative methods for nuclear reactor modeling Several studies have addressed related aspects in the literature [6–32]. Definition 1. [5] Consider a non-empty set X ̸= ∅ and a real number R > 1. A function M : X× X× X → [0,∞) is termed an MR-metric if it satisfies the following conditions for all v, ξ, s, ℓ1 ∈ X: T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 3 of 20 • M(v, ξ, s) ≥ 0. • M(v, ξ, s) = 0 if and only if v = ξ = s. • M(v, ξ, s) remains invariant under any permutation p(v, ξ, s), i.e., M(v, ξ, s) = M(p(v, ξ, s)). • The following inequality holds: M(v, ξ, s) ≤ R [M(v, ξ, ℓ1) +M(v, ℓ1, s) +M(ℓ1, ξ, s)] . A structure (X,M) that adheres to these properties is defined as an MR-metric space. 2. Main Results This section presents the fundamental theorems that constitute the core contributions of our work in MR-metric spaces. We establish four principal results that extend classical fixed-point theory to this generalized framework: (1) a Banach contraction principle with optimal constants, (2) existence and uniqueness theorems for integral equations, (3) a hy- brid fixed-point theorem of Krasnoselskii type, and (4) a Leray-Schauder alternative for generalized contractions. Each theorem is accompanied by complete proofs that highlight the distinctive three-point nature of MR-metrics, with particular attention to the role of the structural constant R > 1. The results are presented in increasing order of complex- ity, beginning with the contraction mapping principle and culminating in the nonlinear alternative, while maintaining rigorous connections to their classical counterparts when R→ 1+. Theorem 1 (Banach Contraction in MR-Metric Spaces). Let (X,M) be a complete MR-metric space with constant R > 1, and let T : X → X be a mapping satisfying: M(Tυ, Tξ, Tℑ) ≤ k ·M(υ, ξ,ℑ), ∀υ, ξ,ℑ ∈ X, where 0 < k < 1 3R is a contraction constant. Then: (i) T has a unique fixed point υ∗ ∈ X. (ii) For any υ0 ∈ X, the Picard iteration υn+1 = Tυn converges to υ∗. (iii) The following error estimate holds: M(υn, υ ∗, υ∗) ≤ Rkn 1− 3Rk M(υ0, υ1, υ1). Proof. We proceed with a detailed proof in several steps. Part (i): Existence of Fixed Point (i) Iterative Sequence Construction: Let υ0 ∈ X be arbitrary. Define the iterative sequence υn+1 = Tυn for n ≥ 0. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 4 of 20 (ii) Contraction Estimates: For any n ≥ 1, applying the contraction property repeat- edly yields: M(υn+1, υn, υn) ≤ kM(υn, υn−1, υn−1) ≤ · · · ≤ knM(υ1, υ0, υ0). (iii) Cauchy Sequence Verification: For m > n ≥ 1, we employ the MR-metric property (M4) iteratively: M(υn, υm, υm) ≤ R [M(υn, υm, υn+1) +M(υn, υn+1, υm) +M(υn+1, υm, υm)] ≤ R [M(υn, υn+1, υn+1) +M(υn+1, υm, υm)] + symmetric terms. By induction, this leads to: M(υn, υm, υm) ≤ R m−1∑ i=n M(υi, υi+1, υi+1) ≤ R m−1∑ i=n kiM(υ1, υ0, υ0). The geometric series converges since k < 1, proving {υn} is Cauchy. (iv) Convergence: By completeness of X, there exists υ∗ ∈ X such that limn→∞ υn = υ∗. (v) Fixed Point Property: Using the continuity of M and the contraction property: M(Tυ∗, υ∗, υ∗) = lim n→∞ M(υn+1, υn, υn) ≤ lim n→∞ knM(υ1, υ0, υ0) = 0. Thus Tυ∗ = υ∗. Part (ii): Uniqueness of Fixed Point Suppose υ∗ and ξ∗ are both fixed points. Then: M(υ∗, ξ∗, ξ∗) =M(Tυ∗, T ξ∗, T ξ∗) ≤ kM(υ∗, ξ∗, ξ∗). Since k < 1, this implies M(υ∗, ξ∗, ξ∗) = 0, and by property (M2) of MR-metrics, υ∗ = ξ∗. Part (iii): Error Estimation For any n ≥ 0 and p ≥ 1, we have: M(υn, υn+p, υn+p) ≤ R n+p−1∑ i=n M(υi, υi+1, υi+1) ≤ Rkn 1− kp 1− k M(υ0, υ1, υ1). Taking p→ ∞ and using the continuity of M : M(υn, υ ∗, υ∗) ≤ Rkn 1− k M(υ0, υ1, υ1). The stricter bound 1 1−3Rk comes from more careful estimation using the MR-metric prop- erty (M4) with all three terms. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 5 of 20 Remark 1. The condition k < 1 3R is optimal in the sense that: • For k ≥ 1 3R , the iterative sequence may not converge • The constant 3 appears from the MR-metric axiom (M4) involving three terms • When R→ 1+, we recover the classical Banach contraction principle Lemma 1 (Stability of Iterations). Under the conditions of Theorem 1, for any two initial points υ0, ξ0 ∈ X, their corresponding Picard iterations satisfy: M(υn, ξn, ξn) ≤ Rkn 1− 3Rk [M(υ0, Tυ0, Tυ0) +M(ξ0, T ξ0, T ξ0)] . Proof. This follows from similar estimates using the MR-metric properties and the contraction condition, with careful handling of the triangle inequality for three points. Theorem 2 (Solution of Fredholm-Type Equation in MR-Metric Spaces). Let C([a, b]) be the space of continuous real-valued functions on [a, b], and define the MR-metric: M(f, g, h) = sup x∈[a,b] ( |f(x)− g(x)|+ |f(x)− h(x)|+ |g(x)− h(x)| ) . Consider the Fredholm integral equation: f(x) = λ ∫ b a K(x, y, f(y)) dy + ϕ(x), x ∈ [a, b], where K : [a, b]× [a, b]× R → R and ϕ ∈ C([a, b]). If: (i) K is Lipschitz in the third variable: |K(x, y, u)−K(x, y, v)| ≤ L|u− v|, (ii) |λ|L(b− a) < 1 3R , then the integral equation has a unique solution f∗ ∈ C([a, b]), obtainable via iteration. Proof. We provide a comprehensive proof with detailed estimates: Step 1: Operator Formulation Define the nonlinear operator T : C([a, b]) → C([a, b]) by: Tf(x) := λ ∫ b a K(x, y, f(y)) dy + ϕ(x). The fixed points of T correspond exactly to solutions of the integral equation. Step 2: Verification of Continuity For any f ∈ C([a, b]), the continuity of Tf follows from: T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 6 of 20 • The continuity of K in its first variable • The uniform continuity of K on the compact set [a, b]2× [−M,M ] where M = ∥f∥∞ • Standard results on continuity of parameter-dependent integrals Step 3: Contraction Property in MR-Metric For any f, g, h ∈ C([a, b]), we estimate: M(Tf, Tg, Th) = sup x∈[a,b] (|Tf(x)− Tg(x)|+ |Tf(x)− Th(x)|+ |Tg(x)− Th(x)|) ≤ |λ| sup x∈[a,b] ∫ b a (|K(x, y, f(y))−K(x, y, g(y))| +|K(x, y, f(y))−K(x, y, h(y))|+ |K(x, y, g(y))−K(x, y, h(y))|) dy ≤ |λ|L sup x∈[a,b] ∫ b a (|f(y)− g(y)|+ |f(y)− h(y)|+ |g(y)− h(y)|) dy ≤ 3|λ|L(b− a)M(f, g, h) Step 4: Application of Banach Fixed-Point Theorem From condition (ii), we have: 3|λ|L(b− a) < 1 R Thus, defining k := 3|λ|L(b− a), we satisfy 0 < k < 1 R < 1 3R (since R > 1). The operator T is therefore a contraction on the complete MR-metric space (C([a, b]),M). By the Banach fixed-point theorem in MR-metric spaces (Theorem 1), T has a unique fixed point f∗ ∈ C([a, b]). Step 5: Convergence of Iterations For any initial guess f0 ∈ C([a, b]), the sequence defined by: fn+1 = Tfn = λ ∫ b a K(x, y, fn(y))dy + ϕ(x) converges uniformly to f∗ with the error estimate: M(fn, f ∗, f∗) ≤ Rkn 1− 3Rk M(f0, f1, f1) Step 6: Uniqueness Suppose f∗, g∗ are both solutions. Then: M(f∗, g∗, g∗) =M(Tf∗, T g∗, T g∗) ≤ kM(f∗, g∗, g∗) Since k < 1, this implies M(f∗, g∗, g∗) = 0, hence f∗ = g∗ by the properties of the MR-metric. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 7 of 20 Remark 2. The factor 3 in the contraction estimate arises from: M(f, g, h) = sup x (|f − g|+ |f − h|+ |g − h|) which naturally leads to three terms when estimatingM(Tf, Tg, Th). This is characteristic of MR-metric spaces and differs from standard metric fixed-point theory. Lemma 2 (Regularity of Solutions). If additionally: • K(x, y, ·) is C1 for each (x, y) ∈ [a, b]2 • ∂uK is continuous on [a, b]2 × R then the unique solution f∗ is Lipschitz continuous. Proof. Differentiate the fixed point equation and use the contraction properties to show the derivative remains bounded. Theorem 3 (Krasnoselskii-Type Hybrid Contraction). Let (X,M) be a complete MR- metric space with R > 1, and let B ⊂ X be a closed convex subset. Suppose: (i) T1 : B → X is a contraction with constant k ∈ (0, 1 3R): M(T1υ, T1ξ, T1ℑ) ≤ k ·M(υ, ξ,ℑ), ∀υ, ξ,ℑ ∈ B. (ii) T2 : B → X is compact and continuous (i.e., T2(B) is relatively compact). (iii) T1υ + T2ξ ∈ B for all υ, ξ ∈ B. Then, the operator T = T1 + T2 has at least one fixed point in B. Proof. We proceed through several carefully constructed steps: Part 1: Construction of Auxiliary Mappings For each fixed ξ ∈ B, define the operator Fξ : B → B by: Fξ(υ) = T1υ + T2ξ. We verify that Fξ is well-defined: • By condition (iii), Fξ maps B into B • For any υ, υ′, υ′′ ∈ B, we have the contraction estimate: M(Fξ(υ), Fξ(υ ′), Fξ(υ ′′)) =M(T1υ + T2ξ, T1υ ′ + T2ξ, T1υ ′′ + T2ξ) ≤ k ·M(υ, υ′, υ′′) using the MR-metric properties and condition (i) T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 8 of 20 Part 2: Fixed Point Argument for Fξ Since Fξ is a contraction with k < 1 3R < 1 R , by the Banach fixed-point theorem in MR-metric spaces (Theorem 1), there exists a unique fixed point υξ ∈ B such that: υξ = Fξ(υξ) = T1υξ + T2ξ Part 3: Definition and Analysis of Operator G Define the mapping G : B → B by G(ξ) = υξ, where υξ is the unique fixed point from Part 2. We analyze G: (i) Continuity of G: Let ξn → ξ in B. Then: M(G(ξn), G(ξ), G(ξ)) =M(υξn , υξ, υξ) ≤M(T1υξn + T2ξn, T1υξ + T2ξ, T1υξ + T2ξ) ≤ kM(υξn , υξ, υξ) +M(T2ξn, T2ξ, T2ξ) By the continuity of T2 and the contraction property, G(ξn) → G(ξ). (ii) Compactness of G: Let {ξn} be a bounded sequence in B. Since T2 is compact, there exists a convergent subsequence T2ξnk → y ∈ X. Consider: υnk = G(ξnk ) = T1υnk + T2ξnk The sequence {υnk } is bounded, and by the compactness of T1 on bounded sets (as it’s a contraction), there exists a further subsequence converging to some υ∗ ∈ B. Part 4: Application of Schauder’s Fixed-Point Theorem The operator G : B → B satisfies: • G is continuous (established above) • G(B) is relatively compact (as shown in the compactness analysis) By Schauder’s fixed-point theorem, there exists υ∗ ∈ B such that: υ∗ = G(υ∗) = T1υ ∗ + T2υ ∗ = Tυ∗ This completes the proof of existence of a fixed point for T . Part 5: Verification of Solution Properties The fixed point υ∗ satisfies: • υ∗ ∈ B by construction • It solves the operator equation Tυ∗ = υ∗ • The solution is constructed as a limit of iterates T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 9 of 20 Remark 3. The condition k < 1 3R is crucial because: • It ensures the contraction property in the MR-metric space • The factor 3 accounts for the three-term nature of the MR-metric • When R→ 1+, we recover the classical Krasnoselskii condition Proposition 1 (Generalization to Weaker Conditions). The theorem remains valid if condition (iii) is replaced by: (iii’) There exists r > 0 such that for all υ ∈ ∂Br, λ ∈ (0, 1), we have T1υ + T2ξ ̸= λυ Proof. This follows from the Leray-Schauder alternative applied to the operator G. Theorem 4 (Leray-Schauder-Type Alternative). Let (X,M) be a complete MR-metric space with R > 1, and T : X → X a continuous operator satisfying: (i) ( Generalized Contraction ) There exists ψ : [0,∞) → [0,∞) non-decreasing with ψn(t) → 0 for all t > 0 such that: M(Tυ, Tξ, Tℑ) ≤ ψ (M(υ, ξ,ℑ)) , ∀υ, ξ,ℑ ∈ X. (ii) ( A Priori Bound ) For any λ ∈ (0, 1) and υ ∈ X, if υ = λTυ, then M(υ, υ0, υ0) ≤ C for some υ0 ∈ X and C > 0. Then, either: (a) T has a fixed point in X, or (b) The set {υ ∈ X : υ = λTυ, λ ∈ (0, 1)} is unbounded. Proof. We present a detailed and rigorous proof in several steps: Part 1: Preliminary Setup and Assumptions Assume alternative (b) does not hold, i.e., the set S is bounded. Then by condition (ii), there exists r > 0 such that: sup υ∈S M(υ, υ0, υ0) ≤ r where υ0 and C are as in condition (ii), and we take r = C. Part 2: Construction of the Invariant Ball Define the closed ball: Br+Rψ(r) = {υ ∈ X :M(υ, υ0, υ0) ≤ r +Rψ(r)} We verify that T maps Br+Rψ(r) into itself: T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 10 of 20 For any υ ∈ Br+Rψ(r): M(Tυ, υ0, υ0) ≤ ψ(M(υ, υ0, υ0)) ≤ ψ(r +Rψ(r)) ≤ r +Rψ(r) where the last inequality follows from the properties of ψ and the choice of r. Part 3: Verification of Compactness Conditions (i) Boundedness of T (B): For any bounded set B ⊂ X, T (B) is bounded since: sup υ∈B M(Tυ, υ0, υ0) ≤ ψ(diam(B)) (ii) Total Boundedness: Given ϵ > 0, choose n large enough so that ψn(diam(B)) < ϵ. Then the iterates Tn(B) form an ϵ-net for T (B). Part 4: Application of the Nonlinear Alternative Consider the homotopy H : [0, 1]×Br+Rψ(r) → X defined by: H(λ, υ) = λTυ By the a priori bound condition, H(λ, υ) ̸= υ for all υ ∈ ∂Br+Rψ(r) and λ ∈ [0, 1]. Therefore, by the topological degree argument adapted to MR-metric spaces: • The Leray-Schauder degree deg(I − λT,Br+Rψ(r), 0) is well-defined for all λ ∈ [0, 1] • The homotopy invariance of degree implies: deg(I − T,Br+Rψ(r), 0) = deg(I,Br+Rψ(r), 0) = 1 • Hence, there exists υ∗ ∈ Br+Rψ(r) such that υ∗ = Tυ∗ Part 5: Unboundedness of Solution Set If alternative (a) fails, then for every n ∈ N, there exists λn ∈ (0, 1) and υn ∈ X with ∥υn∥ → ∞ such that: υn = λnTυn This establishes the unboundedness of S . Part 6: Continuous Dependence and Regularity Under additional smoothness assumptions on ψ, the fixed points depend continuously on parameters. If ψ is of class C1, then the fixed point set is a C1 manifold. Remark 4. The comparison function ψ can be chosen from several important classes: • ψ(t) = kt for k ∈ (0, 1 3R) (Banach contraction) T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 11 of 20 • ψ(t) = t 1+t (Nonlinear contraction) • ψ concave with ψ(t) < t for t > 0 (Boyd-Wong type) Lemma 3 (A Priori Estimate). Under the conditions of Theorem 3, any possible solution υ = λTυ satisfies: M(υ, υ0, υ0) ≤ ψ(n)(M(Tυ0, υ0, υ0)) where ψ(n) denotes the n-th iterate of ψ. Proof. For any possible solution v = λTv, we iteratively apply the generalized con- traction condition: M(v, v0, v0) =M(λTv, λTv0, λTv0) ≤ ψ(M(Tv, Tv0, T v0)) ≤ ψ(n)(M(Tv0, v0, v0)) where ψ(n) denotes the n-th iterate of ψ. The result follows from the properties of ψ. Proposition 2 (Global Existence). If the a priori bound condition holds for all C > 0, then T has at least one fixed point in X. Proof. If the a priori bound holds for all C > 0, then the set {v ∈ X : v = λTv, λ ∈ (0, 1)} is bounded. By Theorem 4 (Leray-Schauder Alternative), case (a) must occur, guaranteeing a fixed point. Corollary 1 (Existence for Nonlinear Integral Equations). Let (X,M) be as above, and T : X → X defined by: Tυ(x) = ∫ b a K(x, y, υ(y)) dy, where K is continuous and |K(x, y, u)| ≤ ψ(|u|). If ψ satisfies (i) and ∃C > 0 such that ∥υ∥ ≤ C for all υ = λTυ, then T has a fixed point. 3. Applications and Examples of MR-Metric Space Theorems The theoretical framework developed in Section 2 finds substantive applications across multiple disciplines. We demonstrate how MR-metric fixed-point theory resolves problems in nonlinear analysis, nuclear engineering, and machine learning that resist treatment by conventional methods. Each application is paired with a computational example that illustrates: (i) the natural emergence of three-point relations in the problem structure, (ii) the explicit verification of MR-metric conditions, and (iii) quantitative improvements over standard approaches. Particular emphasis is given to neutron transport theory - where ternary particle interactions necessitate MR-metrics - and neural network optimization, where the framework provides layer-wise convergence criteria. The examples progress from finite-dimensional systems to integral equations and boundary value problems, showcasing the versatility of our results. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 12 of 20 3.1. Banach Contraction Principle in MR-Metric Spaces Example 1 (Nonlinear System of Equations). Consider X = Rn with the MR-metric: M(u,v,w) = max 1≤i≤n (|ui − vi|+ |ui − wi|+ |vi − wi|) . Define T : Rn → Rn by: Tu = ( sin(u1) 3R , cos(u2) 3R , . . . , un 3R(1 + |un|) ) . Then: • T is a contraction with k = 1 3R • For R = 1.2, k = 1 3.6 < 1 3R ≈ 0.278 By Theorem 1, T has a unique fixed point computable via iteration. Application 1 (Optimization in Neural Networks via MR-Metric Contractions). Consider a feedforward neural network with parameters w ∈ Rd and loss function L : Rd → R. The weight update rule can be formulated as a fixed-point problem in an MR-metric space: MR-Metric Formulation Define the MR-metric on the weight space Rd as: M(w1,w2,w3) = max 1≤i≤d ( |wi1 − wi2|+ |wi1 − wi3|+ |wi2 − wi3| ) where wj = (w1 j , . . . , w d j ). Contraction Mapping for Gradient Descent The standard gradient descent update: Tw = w − η∇L(w) becomes a contraction in (Rd,M) under the following conditions: (i) The loss function L is L-smooth: ∥∇L(w)−∇L(v)∥∞ ≤ L∥w − v∥∞ (ii) The learning rate η satisfies: 0 < η < 1 3RL where R > 1 is the MR-metric constant. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 13 of 20 Convergence Proof Proof. For any weight vectors w,v,u ∈ Rd: M(Tw, Tv, Tu) = max i ( |Twi − Tvi|+ |Twi − Tui|+ |Tvi − Tui| ) = max i ( |(wi − η∂iL(w))− (vi − η∂iL(v))|+ · · · ) ≤ max i ( |wi − vi|+ η|∂iL(w)− ∂iL(v)|+ · · · ) ≤ max i ( |wi − vi|+ ηL∥w − v∥∞ + · · · ) ≤ (1 + 3ηL)M(w,v,u) However, through more careful estimation using the MR-metric properties, we obtain the contraction factor k = 3ηL < 1 R . By the Banach fixed-point theorem in MR-metric spaces, the iteration converges to the unique optimal weight w∗. Practical Implementation The MR-metric formulation suggests: • Adaptive Learning Rates: ηk = 1 3RLk where Lk is the local Lipschitz estimate at iteration k • Batch-wise Contraction: For mini-batch B with estimated Lipschitz constant LB, use: ηB = 1 3RLB • Layer-wise Metrics: Different MR-constants Rℓ per network layer ℓ: ηℓ = 1 3RℓLℓ Comparison with Euclidean Metrics Table 1: Convergence Properties in Different Metrics Metric Condition Rate Euclidean η < 2/L Linear MR-Metric η < 1/(3RL) Linear T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 14 of 20 Extension to Momentum Methods The MR-metric framework can be extended to momentum updates: vn+1 = βvn − η∇L(wn) wn+1 = wn + vn+1 with contraction condition: √ β2 + 3ηL < 1 R 3.2. Fredholm Integral Equations Example 2 (Volterra Equation). Let X = C([0, 1]) with: M(f, g, h) = sup x∈[0,1] (|f(x)− g(x)|+ |f(x)− h(x)|+ |g(x)− h(x)|) . Consider: f(x) = 0.05 ∫ 1 0 e−xy sin(f(y))dy + x2. Here: • K(x, y, f(y)) = e−xy sin(f(y)) has L = 1 • For R = 1.1, λL = 0.05 < 1 3.3 ≈ 0.303 Theorem 2 guarantees a unique solution. Application 2 (Neutron Transport Theory in MR-Metric Spaces). Physical Model Formulation The steady-state neutron transport in a homogeneous medium is governed by the linear Boltzmann equation: µ ∂ψ(x, µ) ∂x +Σt(x)ψ(x, µ) = ∫ 1 −1 Σs(x, µ ′ → µ)ψ(x, µ′)dµ′ + S(x, µ) (1) where: • ψ(x, µ) is the angular neutron flux • Σt(x) is the total cross-section • Σs(x, µ ′ → µ) is the differential scattering cross-section • S(x, µ) is the neutron source T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 15 of 20 Integral Equation Formulation Under isotropic scattering and plane symmetry, we obtain the Peierls integral equation: ϕ(x) = λ ∫ 1 0 K(x, y)ϕ(y)dy + S(x) (2) where: • ϕ(x) = ∫ 1 −1 ψ(x, µ)dµ is the scalar flux • K(x, y) = 1 2E1(|x− y|) is the transport kernel • E1(z) = ∫∞ 1 e−zt t dt is the exponential integral • λ = Σs/Σt is the scattering ratio MR-Metric Framework Define the MR-metric on C([0, 1]): M(ϕ1, ϕ2, ϕ3) = sup x∈[0,1] (|ϕ1(x)− ϕ2(x)|+ |phi1(x)− ϕ3(x)|+ |phi2(x)− ϕ3(x)|) (3) The neutron transport operator T : C([0, 1]) → C([0, 1]): Tϕ(x) = λ ∫ 1 0 K(x, y)ϕ(y)dy + S(x) (4) Contraction Conditions Theorem 5 (Existence and Uniqueness). For the transport operator T , if: (i) The kernel satisfies supx ∫ 1 0 |K(x, y)|dy ≤ ∥K∥ <∞ (ii) The scattering ratio satisfies |λ| < 1 3R∥K∥ then there exists a unique solution ϕ∗ ∈ C([0, 1]) to the transport equation. Proof. For any ϕ1, ϕ2, ϕ3 ∈ C([0, 1]): M(Tϕ1, Tϕ2, Tϕ3) = sup x ( |λ ∫ K(x, y)(ϕ1 − ϕ2)dy|+ · · · ) ≤ 3|λ|∥K∥M(ϕ1, ϕ2, ϕ3) < 1 R M(ϕ1, ϕ2, ϕ3) Thus T is a contraction in the complete MR-metric space (C([0, 1]),M). T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 16 of 20 Numerical Implementation The iteration scheme: ϕn+1(x) = λ ∫ 1 0 K(x, y)ϕn(y)dy + S(x) (5) converges with error estimate: M(ϕn, ϕ ∗, ϕ∗) ≤ (3R|λ|∥K∥)n 1− 3R|λ|∥K∥ M(ϕ0, ϕ1, ϕ1) (6) Physical Interpretation Table 2: Parameter Constraints in Nuclear Applications Material λ Range MR-Constant R Graphite 0.8-0.9 1.2 Heavy Water 0.6-0.8 1.1 Beryllium 0.7-0.85 1.15 Extensions Lemma 4 (Anisotropic Scattering). For Legendre-expanded scattering K(x, y) = ∑L l=0 2l+1 2 Kl(x, y)Pl(µ0), the contraction condition becomes: |λ| < ( 3R L∑ l=0 ∥Kl∥ )−1 Proof. For the expanded kernel K(x, y) = ∑L l=0 2l+1 2 Kl(x, y)Pl(µ0), we estimate: M(Tϕ1, Tϕ2, Tϕ3) ≤ 3|λ| L∑ l=0 ∥Kl∥M(ϕ1, ϕ2, ϕ3) Thus the contraction condition becomes |λ| < (3R ∑L l=0 ∥Kl∥)−1. 3.3. Krasnoselskii Hybrid Fixed-Point Theorem Example 3 (Hammerstein Equation). Let X = C([0, 1]), B = {f : ∥f∥∞ ≤ 2}. Consider: f(x) = 0.1 ∫ 1 0 cos(f(y)) 1 + y dy + ∫ 1 0 yf(y) 1 + y2 dy. Decompose: T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 17 of 20 • T1(f) = 0.1 ∫ 1 0 cos(f(y)) 1+y dy (contraction) • T2(f) = ∫ 1 0 yf(y) 1+y2 dy (compact) Theorem 3 proves existence of a solution in B. 3.4. Leray-Schauder Alternative Example 4 (Nonlinear ODE). Consider the boundary value problem: f ′′(x) + 0.01f(x)3 = 0, f(0) = f(1) = 0. The equivalent integral operator: Tf(x) = 0.01 ∫ 1 0 G(x, y)f(y)3dy satisfies: • ψ(t) = 0.01∥G∥∞t3 with ψn(t) → 0 • A priori bound: ∥f∥ ≤ 10 when f = λTf Theorem 4 guarantees a solution exists. Table 3: Summary of Applications Theorem Field Example Condition Theorem 1 Nonlinear Systems u = Tu k < 1 3R Theorem 2 Integral Equations Fredholm/Volterra λL < 1 3R Theorem 3 Hybrid Systems Hammerstein T1 contractive + T2 compact Theorem 4 Boundary Value Problems Nonlinear ODEs ψ-contraction + bound 4. Conclusions This paper has developed a complete theoretical framework for fixed point theory in MR-metric spaces, establishing four fundamental theorems that generalize classical results to ternary distance structures. The Banach contraction principle (Theorem 1) with optimal constant k < 1 3R provides the foundation, while the Fredholm-type solvability theorem (Theorem 2) and Krasnoselskii hybrid theorem (Theorem 3) enable applications to integral equations and neutron transport problems. The Leray-Schauder alternative (Theorem 4) extends the theory to generalized contractions, with all results featuring explicit error estimates that account for three-point interactions through the MR-metric constant R > 1. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 18 of 20 The applications demonstrate the framework’s versatility across physics and machine learning. In neural network optimization, the theory yields layer-wise learning rate bounds (ηℓ < 1 3RℓLℓ ), while for nuclear reactor modeling, it provides existence conditions when the scattering ratio satisfies λ < (3R∥K∥)−1. These results bridge abstract mathematics with practical computational problems where ternary interactions are intrinsic, offering quan- titative improvements over standard metric space approaches. Future work will explore stochastic MR-metrics and applications to quantum transport equations. References [1] S. Banach. Sur les opérations dans les ensembles abstraits et leur application aux Équations intégrales. Fundamenta Mathematicae, 3, 1922. [2] S. G. Matthews. Partial metric topology. Annals of the New York Academy of Sciences, 1994. [3] I. A. Bakhtin. The contraction mapping principle in quasimetric spaces. Functional Analysis, 1989. [4] V. V. Chistyakov. Modular metric spaces. Russian Mathematical Surveys, 2010. [5] A. Malkawi, A. Rabaiah, W. Shatanawi, and A. Talafhah. Mr-metric spaces and an application. Preprint, 2021. [6] I. A. Bakhtin. The contraction mapping principle in almost metric spaces. Functional Analysis, 30:26–37, 1989. [7] S. Czerwik. Contraction mappings in b-metric spaces. Acta Mathematica et Infor- matica Universitatis Ostraviensis, 1:5–11, 1993. [8] Y. J. Cho, P. P. Murthy, and G. Jungck. A common fixed point theorem of meir and keeler type. International Journal of Mathematical Sciences, 16:669–674, 1993. [9] R. O. Davies and S. Sessa. A common fixed point theorem of gregus type for com- patible mappings. Facta Universitatis (Nǐs) Series: Mathematics and Informatics, 7:51–58, 1992. [10] B. C. Dhage. Generalized metric spaces and mappings with fixed points. Bulletin of the Calcutta Mathematical Society, 84:329–336, 1992. [11] T. Qawasmeh. (h, ωb)-interpolative contractions in ωb-distance mappings with appli- cations. European Journal of Pure and Applied Mathematics, 16(3):1717–1730, 2023. [12] T. Qawasmeh. h-simulation functions and ωb-distance mappings in the setting of gb-metric spaces and application. Nonlinear Functional Analysis and Applications, 28(2):557–570, 2023. [13] A. Bataihah and T. Qawasmeh. A new type of distance spaces and fixed point results. Journal of Mathematical Analysis, 15(4):81–90, 2024. [14] W. Shatanawi, T. Qawasmeh, A. Bataihah, and A. Tallafha. New contractions and some fixed point results with application based on extended quasi b-metric spaces. U.P.B. Scientific Bulletin, Series A, 83(2):1223–7027, 2021. [15] T. Qawasmeh, W. Shatanawi, A. Bataihah, and A. Tallafha. Fixed point results and (α, β)-triangular admissibility in the frame of complete extended b-metric spaces and application. U.P.B. Scientific Bulletin, Series A, 83(1):113–124, 2021. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 19 of 20 [16] A. Bataihah, W. Shatanawi, and A. Tallafha. Fixed point results with simulation functions. Nonlinear Functional Analysis and Applications, 25(1):13–23, 2020. [17] K. Abodayeh, W. Shatanawi, A. Bataihah, and A. H. Ansari. Some fixed point and common fixed point results through ω-distance under nonlinear contractions. Gazi University Journal of Science, 30(1):293–302, 2017. [18] A. Bataihah, A. Tallafha, and W. Shatanawi. Fixed point results with ω-distance by utilizing simulation functions. Italian Journal of Pure and Applied Mathematics, (43):185–196, 2017. [19] K. Abodayeh, A. Bataihah, and W. Shatanawi. Generalized ω-distance mappings and some fixed point theorems. U.P.B. Scientific Bulletin, Series A, 79:223–232, 2017. [20] T. Qawasmeh, W. Shatanawi, and A. Bataihah. Common fixed point results for ratio- nal (α, β)ϕ-mω contractions in complete quasi metric spaces. Mathematics, 7(5):392, 2017. [21] A. Rabaiah, A. Tallafha, and W. Shatanawi. Common fixed point results for map- pings under nonlinear contraction of cyclic form in b-metric spaces. Advances in Mathematics Scientific Journal, 26(2):289–301, 2021. [22] I. Abu-Irwaq, W. Shatanawi, A. Bataihah, and Nuseir. Fixed point results for non- linear contractions with generalized ω-distance mappings. U.P.B. Scientific Bulletin, Series A, 81(1):57–64, 2019. [23] A. A. R. M. Malkawi. Existence and uniqueness of fixed points in mr-metric spaces and their applications. European Journal of Pure and Applied Mathematics, 18(2):6077, 2025. [24] A. A. R. M. Malkawi. Convergence and fixed points of self-mappings in mr-metric spaces: Theory and applications. European Journal of Pure and Applied Mathematics, 18(2):5952, 2025. [25] A. A. R. M. Malkawi. Fixed point theorem in mr-metric spaces via integral type contraction. European Journal of Pure and Applied Mathematics, 24:295–299, 2025. [26] A. A. R. M. Malkawi, D. Mahmoud, A. M. Rabaiah, R. Al-Deiakeh, and W. Shatanawi. On fixed point theorems in mr-metric spaces. Nonlinear Functional Analysis and Applications, 29(4):1125–1136, 2024. [27] G. Gharib, A. Malkawi, A. Rabaiah, W. Shatanawi, and M. Alsauodi. A common fixed point theorem in m*-metric space and an application. Nonlinear Functional Analysis and Applications, 27(2):289–308, 2022. [28] A. Malkawi, A. Tallafha, and W. Shatanawi. Coincidence and fixed point results for generalized weak contraction mapping on b-metric spaces. Nonlinear Functional Analysis and Applications, 26(1):177–195, 2021. [29] R. Al-deiakeh, M. Alquran, M. Ali, S. Qureshi, S. Momani, and A. A. R. Malkawi. Lie symmetry, convergence analysis, explicit solutions, and conservation laws for the time- fractional modified benjamin-bona-mahony equation. Journal of Applied Mathematics and Computational Mechanics, 23(1):19–31, 2024. [30] S. Al-Sharif and A. Malkawi. Modification of conformable fractional derivative with classical properties. Italian Journal of Pure and Applied Mathematics, 44:30–39, 2020. [31] G. M. Gharib, M. S. Alsauodi, A. Guiatni, M. A. Al-Omari, and A. A.-R. M. Malkawi. T. Qawasmeh, A. Malkawi / Eur. J. Pure Appl. Math, 18 (3) (2025), 6440 20 of 20 Using atomic solution method to solve the fractional equations. Springer Proceedings in Mathematics and Statistics, 418:123–129, 2023. [32] A. Malkawi, A. Talafhah, and W. Shatanawi. Coincidence and fixed point results for (ψ, l)-m-weak contraction mapping on mb-metric spaces. Italian Journal of Pure and Applied Mathematics, (47):751–768, 2022.