---
title: Three-Operator Splitting Method
url: https://www.emergentmind.com/topics/three-operator-splitting
type: topic
---

# Three-Operator Splitting Method

A three-operator splitting method is an algorithmic framework designed to solve problems where a target operator (typically monotone, or associated with the first-order optimality condition of a minimization problem) naturally decomposes into a sum of three constituent operators, at least one of which has a smooth structure amenable to explicit treatment. This splitting approach generalizes the classic two-operator methods such as Douglas–Rachford and Forward–Backward schemes, and provides increased modeling flexibility and computational efficiency for structured monotone inclusions, composite convex minimization, and partial differential equations.

## 1. Mathematical Formulation and Fundamental Principles

Three-operator splitting addresses monotone inclusion problems of the form:
\[
\text{Find } x\in \mathcal{H} \text{ such that } 0\in A(x) + B(x) + C(x),
\]
where $A: \mathcal{H} \rightrightarrows \mathcal{H}$ and $B: \mathcal{H} \rightrightarrows \mathcal{H}$ are maximally monotone operators, and $C: \mathcal{H} \to \mathcal{H}$ is monotone and typically $\beta$-cocoercive (i.e., $\langle Cx - Cy, x-y \rangle \ge \beta \|Cx - Cy\|^2$ for some $\beta > 0$) [1504.01032].

This framework encompasses convex minimization problems:
\[
\min_{x\in \mathcal{H}} f(x) + g(x) + h(x),
\]
with $f,g$ proper, closed, convex and $h$ convex with $L$-Lipschitz gradient (so $C=\nabla h$ and $\beta = 1/L$).

## 2. Classical and Algorithmic Schemes

The seminal Davis–Yin three-operator splitting algorithm [1504.01032] is given as:
- Let $\gamma \in (0,2\beta)$ and start with $z^0\in \mathcal{H}$.
- Iterate:
  1. $x_B^k = J_{\gamma B}(z^k)$,
  2. $x_A^k = J_{\gamma A}(2x_B^k - z^k - \gamma C x_B^k)$,
  3. $z^{k+1} = z^k + \lambda_k (x_A^k - x_B^k)$, $\lambda_k\in (0,1/\alpha)$ with $\alpha=2\beta/(4\beta-\gamma)$.

The operator $T := I - J_{\gamma B} + J_{\gamma A}(2J_{\gamma B} - I - \gamma C J_{\gamma B})$ is $\alpha$-averaged for $\gamma\in(0,2\beta)$ and fixed points of $T$ are mapped to zeros of $A+B+C$ via $x^* = J_{\gamma B}(z^*)$ [1504.01032, 1904.11684].

Many variants and extensions exist:
- Forward–Douglas–Rachford splitting, where the prox evaluations and gradient steps are permuted [1704.06948].
- Inertial and momentum-accelerated schemes for accelerated convergence [1904.11684, 2511.14050, 2410.01099].
- Adaptive step-size variants (e.g., ATOS) leveraging local smoothness for automatic step selection [1804.02339].
- Bregman-distance generalizations tuning the geometry of the proximal subproblems [2203.00252].
- Extensions for nonconvex settings using modified envelopes and merit functions [2504.07364, 2006.08951, 2103.04568].

## 3. Convergence Theory and Complexity Results

Under the basic monotonicity and cocoercivity assumptions, three-operator splitting methods admit robust convergence guarantees:
- **Weak convergence:** For the Krasnosel’skiĭ–Mann iteration with suitable relaxation, the sequence $(z^k)$ converges weakly to a fixed point of $T$, mapping to a zero of $A+B+C$ [1504.01032, 1904.11684].
- **Rates:** In the convex case, nonergodic $o(1/\sqrt{k})$ rates for the normed residual and $O(1/k)$ ergodic rates for function value and variational-inequality error can be established [1504.01032, 1610.07830, 1910.04229]. With strong monotonicity (e.g., one strongly monotone operator), linear ($R$-linear) convergence is obtained [1504.01032, 2511.14050].
- **Acceleration:** Allowing variable stepsizes or adapting parameters further improves rates (optimal $O(1/k^2)$ for strongly monotone inclusions under varying $\gamma_k$) [1504.01032].
- **Nonconvex Analysis:** Generalizations under suitable energy decrease and Kurdyka–Łojasiewicz conditions yield subsequential global convergence to critical points [2504.07364, 2006.08951].

A robust Lyapunov/IQC-based control interpretation underpins the rigorous certification of rates and parameter choice, supporting the use of LMIs or SDP to derive optimal step and relaxation choices [1910.04229].

## 4. Structural Flexibility and Special Cases

Three-operator splitting generalizes and recovers numerous classical methods [1504.01032, 1704.06948]:
- **Forward–Backward:** $B=0$ yields the standard FB scheme.
- **Douglas–Rachford:** $C=0$ yields DRS, with splitting between two maximally monotone operators.
- **Forward–Douglas–Rachford:** Suitable for saddle-point problems with additional constraints.
- **3-block ADMM:** The Davis–Yin scheme reproduces three-block ADMM in dual space, clarifying the role of each sub-problem and enabling block-separable algorithms for multi-term convex programs [1807.02067, 2411.00166].

Algorithmic frameworks such as AFBA [1602.08729] and recent modifications allow for bounded linear/skew operators and block-preconditioning, further expanding the admissible problem class to composite minimization, primal-dual, and splitting over arbitrary numbers of operators.

## 5. Practical Enhancements, Adaptivity, and Implementation

Key practical developments include:
- **Automatic or adaptive step-size strategies** (e.g., ATOS, adaptive splitting) which circumvent the need for global smoothness constants and can dramatically speed up convergence compared to fixed-step methods [1804.02339, 2104.05460].
- **Momentum and inertial extrapolation** (both one-step and two-step) accelerate convergence, obviate strict summability conditions, and provide advantages for large-scale and ill-conditioned problems; two-step schemes have shown improvement over one-step inertia in imaging and regression [1904.11684, 2511.14050, 2410.01099].
- **Bregman and variable-metric splitting** use distance-generating functions tailored to problem geometry, facilitating algorithmic preconditioning [2203.00252].

Implementation is modular: only the resolvents/ proximal steps for each operator and explicit evaluations (e.g., gradients for $C$) are needed. In practice, method selection can be guided by the cost profile of the constituent operators; for a highly unbalanced cost, efficiency gains can be realized by adjusting the relative frequency of the sub-steps [2302.08034].

## 6. Applications and Numerical Performance

Three-operator splitting is applied to diverse domains:
- **Large-scale convex optimization:** Machine learning (group lasso, total variation), signal and image processing (nuclear norm regularization, matrix/tensor completion), PDEs, and control (box-constrained LQR).
- **Distributed and stochastic optimization:** Mini-batch and distributed variants support localized data, enabling distributed machine learning under heavy-tailed or adversarial stochastic oracles [2203.04020].
- **Nonconvex optimization:** Adapted algorithms and envelope-based analyses guarantee convergence for structured nonconvex objectives common in modern data science [2504.07364, 2006.08951, 2103.04568].
- **Numerical advantages:** Empirical performance shows iterations and wall-clock time can be reduced by factors ranging from 10–20% (PDE splitting) up to order-of-magnitude gains in high-dimensional or degenerate problems [2302.08034, 1804.02339, 1904.11684].

## 7. Extensions and Future Directions

Ongoing research areas include:
- **Generalization to more than three operators:** Block-coordinate schemes and multi-splitting algorithms—which preserve single-proximal-evaluation structure per operator—underlying large-scale consensus optimization [1704.06948, 1602.08729].
- **Adaptive and line-search step-size selection:** Parameter-free automatic schemes, suitable for situations with unknown operator characteristics [1804.02339, 2203.00252].
- **Robustness to noise and inexactness:** Integration with stochastic and inexact oracles broadens applicability to learning over distributed, federated and noisy environments [2203.04020].
- **Further exploration of momentum, extrapolation, and Bregman geometry:** To improve rate constants and enable efficient large-scale parallel and distributed implementations [2511.14050, 1904.11684].
- **Unified primal-dual and preconditioning perspectives:** Leveraging block operator decompositions for composite problems and deep connections to primal–dual and ADMM-type algorithms [1602.08729, 2411.00166].

The versatility, theoretical guarantees, and empirical performance of three-operator splitting have established it as a foundational methodology in contemporary optimization and computational mathematics.

Source: https://www.emergentmind.com/topics/three-operator-splitting