Fact-checked by Grok 2 weeks ago

Contraction mapping

In mathematics, a contraction mapping, also known as a contractive mapping, is a function f: X \to X defined on a metric space (X, d) such that there exists a constant k \in [0, 1) satisfying d(f(x), f(y)) \leq k \cdot d(x, y) for all x, y \in X, meaning it uniformly reduces distances between points by a factor strictly less than 1.^[1]^[2] The Banach fixed-point theorem, the cornerstone result associated with contraction mappings, asserts that if X is a complete metric space and f is a contraction on X, then f has a unique fixed point x^* \in X where f(x^*) = x^*, and the iterative sequence defined by x_{n+1} = f(x_n) for any initial x_0 \in X converges to x^* at a geometric rate bounded by k^n / (1 - k).^[1]^[3] This theorem guarantees not only existence and uniqueness but also an effective constructive method for approximating the fixed point via successive iterations.^[2] The concept originated with Stefan Banach in his 1922 doctoral dissertation published in Fundamenta Mathematicae, where it was presented as a key tool in functional analysis, though similar ideas appeared earlier in works by mathematicians like Émile Picard for solving differential equations.^[4] Independently generalized by Renato Caccioppoli in 1930, it is sometimes called the Banach–Caccioppoli theorem, and its influence extended Banach's foundational contributions to the field formalized in his 1932 book Théorie des opérations linéaires.^[4] Over the century since its inception, the theorem has seen numerous extensions, such as those by Kolmogorov and Fomin in the 1950s–1960s relaxing the contraction condition to apply to iterates, and further generalizations to non-continuous or probabilistic settings, underscoring its robustness and broad applicability.^[4] Contraction mappings find extensive use across pure and applied mathematics, including proving local existence and uniqueness of solutions to ordinary differential equations via Picard iteration, where the integral operator is shown to be a contraction under Lipschitz conditions on the right-hand side.^[2]^[3] They also underpin numerical methods like Newton's method for root-finding, which converges quadratically under suitable conditions akin to contractions, and appear in integral equations of the second kind, where successive approximations yield solutions.^[3] Beyond analysis, applications extend to computer science and engineering, such as Google's PageRank algorithm modeling web link structures as contractions on graph spaces, and in fractal geometry through iterated function systems that generate self-similar sets via contractive transformations.^[3]^[5] These properties make contraction mappings indispensable for establishing convergence in iterative algorithms and solving nonlinear problems in diverse fields.^[5]

Fundamentals

Definition

In a metric space (X, d), a mapping f: X \to X is called a contraction mapping (or simply a contraction) if there exists a constant k \in [0, 1) such that

d(f(x), f(y)) \leq k \cdot d(x, y)

for all x, y \in X.^[3] This inequality implies that the mapping reduces distances between points by a uniform factor less than 1 across the entire space.^[3] The constant k is known as the Lipschitz constant of the mapping, and the condition k < 1 distinguishes strict contractions from non-strict ones (where k \leq 1), with the former ensuring a uniform shrinkage that is the focus of classical theory.^[3] In contrast to pointwise contractions, where the contraction factor may vary by location or apply only locally, the standard definition requires a single k < 1 to hold globally and uniformly for all pairs of points.^[6] The concept of contraction mappings was introduced by Stefan Banach in 1922, originally in the context of solving functional equations within abstract sets.

Basic Properties

A contraction mapping is a subclass of Lipschitz continuous mappings, distinguished by having a Lipschitz constant k satisfying $0 \le k < 1. This stricter condition ensures that distances between images of points are uniformly reduced relative to their original distances, unlike general Lipschitz mappings where the constant can be any nonnegative number (including values greater than or equal to 1).^[1] Every contraction mapping is continuous. Suppose f: (X, d) \to (X, d) is a contraction with constant k < 1. To verify continuity at an arbitrary point x \in X, fix \varepsilon > 0. Set \delta = \varepsilon / k > 0. Then, for any y \in X with d(y, x) < \delta,

d(f(y), f(x)) \le k \, d(y, x) < k \cdot \frac{\varepsilon}{k} = \varepsilon.

Thus, f is continuous at x. Since the choice of \delta depends only on \varepsilon and not on x, f is in fact uniformly continuous on X.^[1]^[7]

Fixed-Point Theorems

Banach Fixed-Point Theorem

The Banach fixed-point theorem, also known as the contraction mapping theorem or Banach contraction principle, provides a fundamental guarantee of existence and uniqueness for fixed points of certain mappings in metric spaces. Specifically, if (X, d) is a complete metric space and f: X \to X is a contraction mapping with Lipschitz constant k \in [0, 1), then f has a unique fixed point

x^* \in X&#36; satisfying

f(x^) = x^.[8] Moreover, for any initial point x_0 \in X, the sequence of Picard iterates defined by x_{n+1} = f(x_n)forn \geq 0converges tox^*in the metricd$.^[8] A key aspect of the theorem is the explicit rate of convergence provided by the error estimate: d(x_n, x^*) \leq \frac{k^n}{1 - k} \, d(x_0, x_1) for all n \geq 0.^[8] This bound highlights the geometric convergence of the iterates, with the factor k^n ensuring rapid approach to the fixed point as n increases, provided k is sufficiently small. The theorem's strength lies in its constructive nature, allowing the fixed point to be approximated iteratively without additional assumptions beyond the contraction property. The theorem requires two essential conditions: the completeness of the metric space (X, d), which ensures the Cauchy sequence of iterates converges, and the global contraction property, meaning d(f(x), f(y)) \leq k \, d(x, y) holds for all x, y \in X with the same k < 1.^[8] Without completeness, counterexamples exist where contractions fail to have fixed points, such as in the rationals with the standard metric. Similarly, relaxing the global k < 1 to local contractions may yield fixed points but not necessarily uniqueness or guaranteed convergence via iterates. Named after the Polish mathematician Stefan Banach, the theorem was first proved in 1922 as part of his doctoral dissertation, generalizing earlier fixed-point results by Henri Poincaré in 1886 and Joseph Liouville on specific analytic mappings.^[9] Banach's formulation abstracted these ideas to general complete metric spaces, establishing a cornerstone of modern functional analysis and nonlinear problems.^[9]

Proof of the Banach Fixed-Point Theorem

The proof of the Banach fixed-point theorem proceeds by constructing a sequence of iterates and demonstrating its convergence to a unique fixed point in the complete metric space (X, d), where f: X \to X is a contraction mapping with constant k < 1. Begin by selecting an arbitrary initial point x_0 \in X. Define the sequence of Picard iterates recursively as x_{n+1} = f(x_n) for n \geq 0. This sequence is well-defined since f maps X into itself.^[10] To establish convergence, first show that the sequence is Cauchy. For integers m > n \geq 0,

d(x_m, x_n) \leq \sum_{i=n}^{m-1} d(x_{i+1}, x_i) \leq \sum_{i=n}^{m-1} k^i \, d(x_1, x_0) \leq k^n \, d(x_1, x_0) \sum_{i=0}^{\infty} k^i = \frac{k^n}{1 - k} \, d(x_1, x_0),

where the inequality follows from the contraction property d(f(x), f(y)) \leq k \, d(x, y) applied successively, and the final bound uses the geometric series sum \sum_{i=0}^{\infty} k^i = 1/(1-k) since |k| < 1. As n \to \infty, the right-hand side tends to 0 independently of m > n, so \{x_n\} is a Cauchy sequence.^[10] Since X is complete, the Cauchy sequence \{x_n\} converges to some limit x^* \in X. To verify that x^* is a fixed point, note that f is continuous as a contraction (with d(f(x), f(y)) \leq k \, d(x, y) implying uniform continuity). Thus, taking the limit in x_{n+1} = f(x_n) yields x^* = f(x^*).^[10] For uniqueness, suppose y^* \in X is another fixed point, so f(y^*) = y^*. Then d(x^*, y^*) = d(f(x^*), f(y^*)) \leq k \, d(x^*, y^*). Since k < 1, it follows that (1 - k) d(x^*, y^*) \leq 0, implying d(x^*, y^*) = 0 and hence x^* = y^*. This also shows f is injective.^[10] The convergence rate of the iterates to the fixed point is given by d(x_n, x^*) \leq \frac{k^n}{1 - k} \, d(x_1, x_0), derived similarly by bounding the tail of the geometric series from the Cauchy estimate and using d(x^*, x_m) \leq \sum_{i=m}^{\infty} d(x_{i+1}, x_i) \leq \frac{k^m}{1 - k} \, d(x_1, x_0) for any m, then letting m \to \infty.^[10]

Examples and Applications

Metric Space Examples

A classic example of a contraction mapping occurs in the complete metric space (\mathbb{R}, d), where d(x, y) = |x - y|, with the function f(x) = x/2.^[1]^[3] This mapping satisfies the contraction condition with constant k = 1/2 < 1, since |f(x) - f(y)| = |(x - y)/2| = (1/2)|x - y| for all x, y \in \mathbb{R}.^[1] The unique fixed point is x = 0, solved from x = x/2, implying x/2 = 0.^[3] Starting from any initial x_0 \in \mathbb{R}, the iterates x_{n+1} = f(x_n) converge geometrically to 0 at rate k = 1/2, as the error satisfies |x_n - 0| \leq (1/2)^n |x_0|.^[1] Another illustrative example is the function f(x) = \cos x on the closed interval [-1, 1] subset of \mathbb{R}, equipped with the standard metric d(x, y) = |x - y|.^[3]^[11] Here, f maps [-1, 1] into itself, since \cos x \in [-1, 1] for x \in [-1, 1].^[3] It is a contraction with constant k = \sin 1 \approx 0.8415 < 1, derived from the mean value theorem: |f(x) - f(y)| = |-\sin c| \cdot |x - y| for some c between x and y, and \sup_{[-1,1]} |\sin c| = \sin 1.^[11] The equation x = \cos x thus has a unique solution in [-1, 1], which has no closed-form expression but can be approximated iteratively, converging to approximately 0.739085.^[3] In the space of bounded continuous functions C[0, 1] with the supremum norm \|y\|_\infty = \sup_{x \in [0,1]} |y(x)|, consider the integral equation y(x) = g(x) + \int_0^1 K(x, t) y^2(t) \, dt, where g \in C[0, 1], K \in C([0, 1] \times [0, 1]), and M = \max_{(x,t) \in [0,1]^2} |K(x, t)|.^[12] Define the operator T(y)(x) = g(x) + \int_0^1 K(x, t) y^2(t) \, dt; under the condition \|g\|_\infty < 1/(8M), T maps the closed ball of radius r = 1/(4M) into itself and is a contraction with constant k = 2Mr < 1.^[12] The unique fixed point in this ball solves the equation, with no explicit closed form in general, but the contraction ensures iterative convergence in the sup norm.^[12]

Real-World Applications

Contraction mappings play a pivotal role in numerical analysis through the method of successive approximations, which iteratively solves nonlinear equations by finding fixed points of a mapping g where x = g(x). For root-finding problems like f(x) = 0, the iteration x_{k+1} = g(x_k) converges to the root if g is a contraction on a complete metric space, ensuring linear convergence with rate determined by the contraction constant q < 1.^[3] In particular, Newton's method, defined by x_{k+1} = x_k - f(x_k)/f'(x_k), can be analyzed as a fixed-point iteration under contraction conditions; near a simple root where f'(x^*) \neq 0, the associated mapping exhibits contraction properties that guarantee local convergence.^[13] A specific case is Newton's method for solving f(x) = 0, which exhibits quadratic convergence near a simple root x^* where f(x^*) = 0 and f'(x^*) \neq 0, assuming f is twice continuously differentiable in a neighborhood of x^*, with the initial guess sufficiently close to x^*. The error satisfies e_{k+1} \approx \frac{|f''(x^*)|}{2 |f'(x^*)|} e_k^2.^[14] The simplified variant, using a fixed initial Jacobian A \approx f'(x^{(0)}) in x_{k+1} = x_k - A^{-1} f(x_k), converges linearly if the mapping is a contraction with constant q < 1 in a ball around the solution, provided second derivatives are bounded and the initial guess is sufficiently close.^[15] In the theory of differential equations, contraction mappings underpin the Picard-Lindelöf theorem, which establishes local existence and uniqueness for solutions to initial value problems y' = f(x, y), y(x_0) = y_0. The theorem considers the integral operator T[y](x) = y_0 + \int_{x_0}^x f(t, y(t)) \, dt on the space of continuous functions over a closed interval [x_0 - h, x_0 + h], where h > 0 is small enough that T becomes a contraction with constant k = L h < 1, with L bounding |\partial f / \partial y|.^[16] The Picard iteration y_{n+1} = T[y_n], starting from y_0(x) = y_0, then converges uniformly to the unique solution in this ball.^[16] In economics, contraction mappings facilitate fixed-point iterations to compute equilibrium prices in general equilibrium models, such as those involving dynamic principal-agent problems with continuous choice sets. By showing that the Bellman operator for the principal's value function is a contraction under discounting and compactness assumptions, the theorem ensures a unique equilibrium policy and value function, computable via value function iteration.^[17] For instance, in one-to-one matching models with linear transferable utility, the system of fixed-point equations for equilibrium transfers forms a contraction mapping, guaranteeing unique transfers and stable outcomes.^[18] In computer science, contraction mappings ensure convergence in algorithms like PageRank, which computes webpage importance as the fixed point of the iteration \rho = \alpha P \rho + (1 - \alpha) v, where P is the transition matrix and \alpha < 1 makes the mapping a contraction under the 1-norm, yielding a unique steady-state ranking vector via power iteration.^[19] Variants of k-means clustering, such as contraction clustering algorithms like RASTER and S-RASTER for data streams, leverage contraction principles by progressively shrinking the data space through grid scaling and density-based projections, enabling single-pass convergence to clusters without the local optima issues of standard k-means.^[20] In fractal geometry, an iterated function system (IFS) consists of a finite collection of contraction mappings \{f_1, \dots, f_n\} on a complete metric space, such as the space \mathbb{H} of nonempty compact subsets equipped with the Hausdorff metric. The associated Hutchinson operator F(A) = \bigcup_{i=1}^n f_i(A) is a contraction on \mathbb{H} with constant k = \max_i k_i < 1, where k_i is the Lipschitz constant of f_i. The Banach fixed-point theorem guarantees a unique nonempty compact attractor K \subseteq \mathbb{H} satisfying F(K) = K, which generates self-similar fractals such as the Sierpinski triangle or Barnsley fern.^[21]

Generalizations and Variants

Non-Expansive Mappings

A non-expansive mapping, also known as a 1-Lipschitz mapping, is a function f: X \to X defined on a metric space (X, d) satisfying d(f(x), f(y)) \leq d(x, y) for all x, y \in X.^[22] This condition represents a relaxation of the strict contraction mapping, where the Lipschitz constant k < 1, allowing the distance between images to remain at most the original distance without necessary reduction.^[23] Non-expansive mappings inherit several beneficial properties from their Lipschitz nature. They are uniformly continuous, ensuring that small changes in input produce bounded changes in output relative to the metric. However, unlike strict contractions, non-expansive mappings do not guarantee a unique fixed point in general complete metric spaces; for instance, the identity mapping f(x) = x is non-expansive on any metric space but admits fixed points at every point.^[24] Fixed-point existence for non-expansive mappings requires additional structural assumptions. In finite-dimensional Euclidean spaces, Brouwer's fixed-point theorem ensures a fixed point for any continuous self-mapping of a compact convex set, and since non-expansiveness implies continuity, the result applies directly. For set-valued non-expansive mappings—where the Hausdorff distance between images satisfies a similar inequality—Kakutani's fixed-point theorem guarantees existence when the mapping is upper semicontinuous with nonempty, closed, convex values on a compact convex subset of \mathbb{R}^n. In infinite-dimensional settings, such as Hilbert spaces, the Browder–Göhde–Kirk theorem establishes fixed-point existence for non-expansive mappings on nonempty, closed, bounded, convex subsets, though uniqueness remains absent without further conditions.^[25] Every contraction mapping is non-expansive, but the converse fails, highlighting the broader applicability of this class at the cost of weakened convergence guarantees.^[26]

Firmly Non-Expansive Mappings

In a Hilbert space H, a mapping f: H \to H is firmly nonexpansive if, for all x, y \in H,

\|f(x) - f(y)\|^2 + \|(I - f)(x) - (I - f)(y)\|^2 \leq \|x - y\|^2.

This condition implies that f is nonexpansive, as the first term on the left-hand side is bounded above by the right-hand side. An equivalent formulation is that f satisfies \langle f(x) - f(y), x - y \rangle \geq \|f(x) - f(y)\|^2 for all x, y \in H, which aligns with monotonicity properties in inner product spaces. Firmly nonexpansive mappings are precisely the \frac{1}{2}-averaged operators, meaning f = (1 - \alpha)I + \alpha T for some nonexpansive T and \alpha = \frac{1}{2}. Equivalently, the operator $2f - I is nonexpansive. This characterization extends to the difference operator I - f, which satisfies a 2-Lipschitz condition in certain contexts, reinforcing the mapping's stability. In fixed-point theory, firmly nonexpansive mappings, being a subclass of nonexpansive mappings, admit fixed points on nonempty, compact, convex subsets of Hilbert spaces by the Browder-Göhde-Kirk theorem. They are particularly amenable to iterative methods, such as the Krasnoselskii-Mann iteration x_{n+1} = (1 - \lambda_n) x_n + \lambda_n f(x_n) with \lambda_n \in (0,1), which converges weakly to a fixed point under standard conditions on the sequence \{\lambda_n\}.^[27] This iteration leverages the averaged nature of firmly nonexpansive operators for improved convergence behavior compared to general nonexpansive cases. Firmly nonexpansive mappings play a central role in the theory of monotone operators, where the resolvent J_A = (I + A)^{-1} of a maximal monotone operator A: H \rightrightarrows H is firmly nonexpansive. This connection is foundational in solving variational inequalities, as the fixed points of such resolvents correspond to zeros of A. In optimization, they underpin proximal algorithms for minimizing convex functions, where the proximity operator \prox_{\gamma g}(x) = \argmin_y \left( g(y) + \frac{1}{2\gamma} \|y - x\|^2 \right) is firmly nonexpansive for \gamma > 0. These properties ensure reliable convergence in methods addressing monotone inclusions and related problems.^[28]

Subcontraction Maps

A subcontraction mapping on a metric space (X, d) is defined such that for every point x \in X, there exists a constant k_x \in [0, 1) satisfying d(f(x), f(y)) \leq k_x \, d(x, y) for all y in some neighborhood of x.^[29] In the global variant, the inequality holds for all y \in X with the point-dependent constant k_x.^[29] This notion, also termed a pointwise contraction, relaxes the uniform contraction condition by allowing the contraction factor to vary by location, accommodating mappings that shrink distances unevenly across the space. Unlike the classical contraction mapping, which requires a single uniform constant k < 1 for all pairs of points, subcontractions permit the constants k_x to approach 1 as x varies, facilitating local or partial contraction analysis in spaces where global uniformity fails.^[6] This flexibility is particularly valuable for investigating fixed points of mappings that exhibit contraction behavior only in restricted regions or with point-specific rates. Fixed-point results for subcontractions include Edelstein's theorem, which applies to compact metric spaces: a continuous self-mapping f satisfying d(f(x), f(y)) < d(x, y) for all x \neq y has a unique fixed point, interpretable as a strict subcontraction without an explicit uniform bound. For complete metric spaces, Caristi's theorem extends this by guaranteeing a fixed point for mappings where d(x, f(x)) \leq \psi(x) - \psi(f(x)) for a lower semicontinuous function \psi: X \to \mathbb{R} with closed bounded sublevel sets \{z \in X : \psi(z) \leq \alpha\}, encompassing subcontractions via appropriate choices of \psi. These theorems ensure existence without requiring global uniformity, though uniqueness may depend on additional path-connectedness or compactness assumptions.^[6] An illustrative example arises in non-complete metric spaces, such as the open unit interval (0, 1) with the Euclidean metric, where a mapping may possess a global Lipschitz constant k \geq 1 but admit local contraction constants k_x < 1 at each interior point. For a differentiable f: (0,1) \to (0,1) with |f'(x)| < 1 for all x \in (0,1) yet \sup_{x \in (0,1)} |f'(x)| = 1 (approached near the endpoints), the mapping fails global contraction due to the supremum reaching 1 and the space's incompleteness, but the pointwise-local property supports fixed-point analysis under supplementary conditions like those in Caristi's framework.

Extensions to Specific Spaces

Contractions in Locally Convex Spaces

In locally convex topological vector spaces, the concept of a contraction mapping is adapted to the absence of a single global norm by leveraging the underlying family of seminorms or the uniform structure induced by balanced convex neighborhoods of the origin. A mapping T: X \to X on a locally convex space X is defined as contractive if there exists h \in (0,1) such that for every balanced convex neighborhood U of the origin and all x, y \in X, if x - y \in tU for some t > 0, then Tx - Ty \in h t U. ^[30] This condition ensures that T uniformly scales distances in a topological sense across the generating neighborhoods, generalizing the metric contraction while respecting the locally convex topology. More generally, a (\psi, \phi)-contractive mapping satisfies a similar inclusion with continuous, strictly increasing functions \psi, \phi: [0, \infty) \to [0, \infty) where \psi(0) = \phi(0) and \psi(t) < \phi(t) for t > 0, allowing for broader classes of mappings that still contract neighborhoods asymptotically. ^[30] Fixed-point theorems in this setting extend the Banach fixed-point theorem to non-metric topologies, often requiring sequential completeness or additional structural assumptions. For instance, in a complete locally convex space, a (\psi, \phi)-contractive mapping admits a unique fixed point, with the iterative sequence T^n x_0 converging to it for any initial x_0 \in X. ^[30] Michael's selection theorem plays a key role for set-valued contractions or lower hemicontinuous maps with convex values, guaranteeing a continuous selection that reduces the problem to a single-valued contraction, thereby ensuring fixed points in paracompact domains mapping to Banach spaces. ^[31] Similarly, the Eilenberg-Montgomery fixed-point theorem applies to multi-valued contractions in uniform spaces, asserting the existence of fixed points for upper semicontinuous maps with acyclic values on absolute neighborhood retracts, facilitating analysis in topological dynamics. ^[32] The lack of a global metric in locally convex spaces introduces challenges for convergence, necessitating tools like asymptotic centers—points minimizing limits of seminorm distances in iterative sequences—or absorbing sets, which ensure iterates remain bounded within convex, topology-generating neighborhoods. ^[33] For asymptotically nonexpansive mappings, fixed points exist in weakly compact convex subsets via asymptotic center methods, addressing non-contractive behaviors that approximate contractions over iterations. ^[34] These techniques are essential for proving stability without a uniform distance function. The development of contraction mappings in locally convex spaces emerged in the mid-20th century, building on foundational work in topological fixed-point theory during the 1950s and 1960s to handle infinite-dimensional settings in dynamics and analysis. ^[35] Seminal contributions, such as Krasnoselskii's extensions for nonlinear contractions and Michael's selections, addressed uniform structures beyond norms, influencing applications in topological vector spaces. ^[35]

Contractions in Normed Spaces

In a normed linear space (X, \|\cdot\|), a mapping f: X \to X is called a contraction if there exists a constant k with $0 \leq k < 1 such that

\|f(x) - f(y)\| \leq k \|x - y\|

for all x, y \in X.^[36] This definition leverages the metric d(x, y) = \|x - y\| induced by the norm, specializing the general notion of contractions in metric spaces to the structured setting of normed spaces.^[36] Unlike arbitrary metrics, the norm-induced metric preserves the vector space operations, which facilitates analysis of mappings that respect linearity or affinity. When f is a bounded linear operator T: X \to X, the contraction condition simplifies to the operator norm satisfying \|T\| < 1.^[37] In this case, the spectral radius \rho(T) obeys \rho(T) \leq \|T\| < 1 by the spectral radius formula \rho(T) = \lim_{n \to \infty} \|T^n\|^{1/n}.^[38] Consequently, the resolvent (I - T)^{-1} exists as a bounded linear operator and equals the Neumann series

(I - T)^{-1} = \sum_{n=0}^\infty T^n,

which converges absolutely in the operator norm.^[38] In complete normed spaces (Banach spaces), fixed-point iteration x_{n+1} = f(x_n) for a contraction f converges to the unique fixed point in the norm topology.^[15] This applies particularly to solving linear systems Ax = b, rewritten as x = (I - A)x + b; if \|A\| < 1, then f(x) = (I - A)x + b is a contraction, and the iteration converges to the solution x = (I - A)^{-1}b.^[15] The vector space structure distinguishes contractions in normed spaces from those in general metric spaces, as it admits affine contractions of the form f(x) = Tx + c where T is linear with \|T\| < 1 and c \in X; such mappings satisfy the contraction inequality with the same constant as T.^[39] This affinity enables algebraic manipulations and spectral insights unavailable in purely metric settings. In broader locally convex spaces, contractions can be defined via families of seminorms, extending these properties topologically.^[38]

References

[1]
[PDF] Chapter 3: The Contraction Mapping Theorem - UC Davis Math
X is a con- traction mapping, or contraction, if there exists a constant c, with 0 < c < 1, such. ...
[2]
[PDF] Contraction mapping theorem and ODEs - Stanford University
The contraction mapping theorem concerns maps f : X → X, (X, d) a metric space, and their fixed points. A point x is a fixed point of f if f(x) = x, ...
[3]
[PDF] THE CONTRACTION MAPPING THEOREM 1. Introduction Let f
Introduction. Let f : X → X be a mapping from a set X to itself. We call a point x ∈ X a fixed point of f if f(x) = x. For example, if [a, b] is a closed ...
[4]
The Banach Fixed Point Theorem: selected topics from its hundred ...
Jul 9, 2024 · Among them there was the theorem known today as the Banach Fixed Point Theorem or the Banach Contraction Principle. It is one of the most famous ...
[5]
Contraction Mapping - an overview | ScienceDirect Topics
Contraction mapping is defined as a mapping S from a space into itself such that the distance between the images of any two functions f and g under S is ...
[6]
On fixed points of locally and pointwise contracting maps
May 15, 2016 · We prove that the following self-mappings must have unique fixed points: pointwise contractive (PC) maps on compact rectifiably path connected ...
[7]
[PDF] metric spaces and the contraction mapping principle - UChicago Math
Remark 3.4. Contraction mappings are uniformly continuous, as can be easily ver- ified. Definition 3.5. Let K be a set and f : K −→ K.Missing: injectivity | Show results with:injectivity
[8]
[PDF] Sur les opérations dans les ensembles abstraits et leur application ...
Sur les opérations dans les ensembles abstraits et leur application aux équations intégrales* publié dans Fund. Math. 3 (1922), p. 133–181. Introduction. L ...
[9]
[PDF] The Banach Fixed Point Theorem: selected topics from its hundred ...
Recall that a mapping T is a selfmapping of X if T acts from X into X. In 2007 Richard S. Palais presented another proof of the Banach Contraction Principle (in.<|control11|><|separator|>
[10]
Sur les opérations dans les ensembles abstraits et leur application ...
Sur les opérations dans les ensembles abstraits et leur application aux équations intégrales. Stefan Banach · Fundamenta Mathematicae (1922) ... pdf.png Full (PDF) ...
[11]
[PDF] Contraction Mappings Consider the equation x = cosx . If we plot the ...
If we plot the graphs of y = cosx and y = x, we see that they intersect at a unique point for x ∼ 0.7. This point is called a fixed point of the function cosx.
[12]
None
### Summary of Basic Properties of Contraction Mappings from the Document
[13]
[PDF] Convergence of Newton's Method for a Single Real Equation
Newton's method for finding the zeroes of a single real function is investigated in some detail. Convergence is generally checked using the Contraction Mapping ...
[14]
[PDF] 1 Fixed Point Iteration and Contraction Mapping Theorem
The Contraction Mapping Theorem states that if g(x) is a contraction on a closed set D, then there exists a unique x* in D where g(x*) = x* and iterates ...
[15]
[PDF] Picard's Existence and Uniqueness Theorem
One of the most important theorems in Ordinary Differential Equations is Picard's. Existence and Uniqueness Theorem for first-order ordinary differential ...
[16]
[PDF] Contraction mapping theorem and computational treatment - EconStor
We consider discrete-time dynamic principal–agent problems with continuous choice sets and potentially multiple agents. We prove the existence of a unique.
[17]
Note on solving one-to-one matching models with linear transferable ...
Sep 9, 2024 · We derive a system of fixed-point equations for the equilibrium transfers in a class of one-to-one matching models with linear transferable utility.
[18]
[PDF] Fixed-Point Centrality for Networks - arXiv
Sep 15, 2022 · PageRank centrality is a contraction under vector 1-norm from Rn to Rn. The existence and uniqueness of fixed- point features for these two ...
[19]
S-RASTER: contraction clustering for evolving data streams
Aug 13, 2020 · Contraction Clustering (RASTER) is a single-pass algorithm for density-based clustering of 2D data. It can process arbitrary amounts of data ...
[20]
Fixed Point Theorems for Nonexpansive Type Mappings in Banach ...
A mapping T : Y → Y is called nonexpansive if ∥ T x − T y ∥ ≤ ∥ x − y ∥ and a point z ∈ Y is a fixed point of T if T z = z . It is well-known that a ...
[21]
Metric fixed point theory for nonexpansive mappings defined on ...
Jul 22, 2014 · It is standard practice in metric fixed point theory to reduce fixed point questions for mappings defined on unbounded sets to the bounded ...
[22]
NONEXPANSIVE NONLINEAR OPERATORS IN A BANACH SPACE ...
PROCEEDINGS, 53, 1100-1103 (1965). 2 Browder, F. E., "Fixed-point theorems for noncompact mappings in Hilbert space," these. PROCEEDINGS, 53, 1272-1276 (1965).
[23]
[PDF] Lipschitz Maps in Metric Spaces - eScholarship
Bi-Lipschitz mappings are precisely those that preserve distances, up to a fixed constant factor. If there is a bi-Lipschitz mapping from a space X onto a space ...
[24]
Fixed Points and Common Fixed Points for Orbit-Nonexpansive ...
Mar 31, 2023 · Banach. Easy examples show that nonexpansive self-mappings defined on a complete metric space may fail to have a fixed point.
[25]
Fixed point property for nonexpansive mappings on large classes in ...
In 1965, Browder [5] proved that every Hilbert space has a property satisfying that every nonexpansive mapping defined on any closed, bounded, and convex (cbc) ...
[26]
[PDF] Maximum Entropy-type Methods, Projections, and (Non ... - CARMA
Further, f is firmly nonexpansive if and only if 2f-I is nonexpansive. LEMMA ... To show that T is firmly nonexpansive, 2T-I is shown to be nonexpansive.
[27]
Convergence of Krasnoselskii-Mann iterations of nonexpansive ...
In this paper, we show that a generic nonexpansive operator on a closed and convex, but not necessarily bounded, subset of a hyperbolic space has a unique ...
[28]
[PDF] SOLVING MONOTONE INCLUSIONS VIA COMPOSITIONS OF ...
The basic iterative scheme under consideration involves nonstationary compositions of perturbed averaged nonexpansive operators. The analysis covers proximal ...
[29]
[PDF] 718790.pdf - EMIS
ABSTRACT. Let S be a subset of a metric space. (X,d) and T: S + X be a mapping. In this paper, we define the notion of lower directional increment.<|separator|>
[30]
[PDF] Generalized contraction mapping principle in locally convex ...
The purpose of this paper is to present the concept of contraction mapping in a locally convex topological vector spaces and to prove the generalized ...
[31]
Applications of Michael's selection theorems to fixed point theory
Applying some of Ernest Michael's selection theorems, from recent fixed point theorems on u.s.c. multimaps, we deduce generalizations of the classical ...Missing: contractions | Show results with:contractions
[32]
Fixed Point Theorems for Multi-Valued Transformations - jstor
FIXED POINT THEOREMS FOR MULTI-VALUED. TRANSFORMATIONS.*. By SAMUEL EILENBERG and DEANE MONTGOMERY. 1. Introduction. Recently there have been several ...
[33]
An Asymptotic Fixed Point Theorem for a Locally Convex Space - jstor
with weighted norm and then apply Horn's asymptotic fixed point theorem. One of the points of this note is that Horn's fixed point theorem may be extended so.Missing: centers | Show results with:centers
[34]
[PDF] FIXED POINTS AND THEIR APPROXIMATIONS FOR ... - EMIS
for asymptotically nonexpansive mappings in locally convex spaces. ... 2. FIXED POINTS ... fixed point for nonexpansive self-mapping in a locally convex space.
[35]
ON A FIXED POINT THEOREM OF KRASNOSELSKII FOR ...
In this paper, we consider mappings defined on a subset S of a locally convex vector space E with values in E (not necessarily 5) and satisfy a certain ...<|control11|><|separator|>
[36]
[PDF] THE CONTRACTION MAPPING PRINCIPLE AND SOME ...
May 13, 2009 · It is an iteration scheme, whose convergence may easily be demonstrated by means of the contraction mapping principle. Many other numerical ...<|control11|><|separator|>
[37]
[PDF] Introductory Functional Analysis with Applications
The book is elementary. A background in under- graduate mathematics, in particular, linear algebra and ordinary cal- culus, is sufficient as a prerequisite.
[38]
Spectral Radii of Bounded Operators on Topological Vector Spaces
Apr 8, 2000 · We show that the Gelfand formula for spectral radius and Neumann series can still be naturally interpreted for operators on topological vector ...Missing: contraction | Show results with:contraction
[39]
[PDF] Maps which have fixed points for any Banach space
We examine a class of maps which have fixed points for all Banach spaces. Included in the class are affine mappings and Banach contractions. The emphasis is ...