Mapping the Fitness Landscape: A Structure-Guided Approach to Multi-Modal Optimization
Meng Xiang, Pei Yan · 2026
Read Mapping the Fitness Landscape: A Structure-Guided Approach to Multi-Modal Optimization by Meng Xiang, Pei Yan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-18351.txt.
Introduction
Meng Xiang and Pei Yan tackle multimodal optimization, where the goal is to find many optima rather than one best answer. Their framework, Chaotic Landscape-Decoding Evolution, maps peak-and-basin structure with a neighbor graph and persistence-guided basin growing, then refines each basin separately. On twenty CEC2013 functions and standard multi-objective suites, the 2026 method keeps strong peak coverage and competitive accuracy.
Catalog record
| CREATOR | Meng Xiang, Pei Yan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57214 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.18351
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Mapping the Fitness Landscape: A Structure-Guided Approach to Multi-Modal Optimization Meng Xiang, Pei Yan Abstract . Multimodal optimization requires finding many optima rather than merely keeping a diverse population. Yet most niching-based evolutionary algorithms rely on distances or density estimators without explicitly recovering the underlying peak–basin organization in the decision space, which can lead to pseudo-multimodality: many distinct individuals ultimately collapse into only a few basins. We introduce Chaotic Landscape-Decoding Evolution (CLDE), a decision-space-centric framework that turns multimodal search into a closed loop of decode–value–allocate–refine. CLDE injects controlled global exploration via a logistic chaotic map with a decaying step size, then builds a k k -nearest-neighbor graph on a decoding canvas and performs persistence-guided basin growing that merges peaks only when they are not separated by deep valleys. An adaptive persistence threshold continuously tunes the decoding resolution online to avoid over-fragmentation and over-merging. Guided by the decoded structure, CLDE carries out basin-wise selection and refinement to improve solution quality while preserving basin coverage. We instantiate CLDE as CLDE-S and CLDE-M for single- and multi-objective multimodal optimization. Experiments on 20 CEC2013 functions show that CLDE-S achieves strong peak ratio under the same evaluation budget, while on DTLZ and MMMOP suites CLDE-M attains competitive IGD/IGDx, with pronounced gains on strongly multimodal problems. 1. Introduction In fields such as robotics, drug discovery, and engineering design, multimodal optimization problems (MMOPs) aim to discover and maintain a diverse set of high-quality yet structurally distinct solutions, rather than converging to a single optimum. Formally, a single-objective MMOP (SOMMOP) seeks a set of solutions that are well-separated in the decision space, while each solution attains sufficiently good objective quality: (1) \min_{i\neq j}\,\lVert x^{(i)}-x^{(j)}\rVert_{2}\ \geq\ \varepsilon_{\mathrm{dec}},\qquad f(x^{(i)})\leq f^{\star}+\varepsilon_{\mathrm{obj}},\ \ \forall i. On benchmarks with known optima (e.g., CEC2013), this goal is commonly instantiated by peak identification under a prescribed decision-space accuracy threshold and summarized by the peak ratio (PR). This requirement extends to multi-objective MMOPs (MOMMOPs), where one must additionally manage Pareto trade-offs for a vector objective function \mathbf{f}(x) : (2) \min_{i\neq j}\ \lVert x^{(i)}-x^{(j)}\rVert_{2}\ \geq\ \varepsilon_{\mathrm{dec}},\qquad\max_{i\neq j}\ \big\lVert\mathbf{f}(x^{(i)})-\mathbf{f}(x^{(j)})\big\rVert_{2}\ \leq\ \delta_{\mathrm{obj}}. A central difficulty in MMOPs is the information gap between the decision space and the objective space. Since selection pressure is driven by objective values, an optimizer can be blind to decision-space topology: structurally distinct solutions may look redundant in objective space, leading to over-exploitation of a few dominant basins and the loss of important modes. Many niching and diversity-preserving methods alleviate mode collapse, but they typically act on individuals and lack explicit mechanisms to identify stable decision-space regions and to quantify which regions deserve more search effort. As a result, evaluations are wasted on redundant or stagnant areas, while minor but promising basins remain under-explored. To address this root cause, we propose CLDE, a region-centric framework that turns multimodal search into a closed-loop process over decision-space regions (Figure. 1 ). CLDE iterates three steps: it decodes stable regions from sampled solutions, assigns each region a saliency score and allocates budgets accordingly, and performs region-aware variation with controlled chaotic perturbations. This cycle bridges the structural information gap by making basin structure explicit, using region valuation for principled budget control, and enabling cross-basin transitions without sacrificing local refinement. Figure 1 . Overview of the CLDE evolutionary cycle with three synergistic modules. (A) Persistence-guided region discovery : build a decision-space neighborhood graph (e.g., kNN) and decode stable basins/regions by merging shallow structures. (B) Region valuation & adaptive budget allocation : compute region values and allocate search budgets proportionally to focus evaluations on informative regions. (C) Region-aware variation with chaotic perturbation : refine solutions locally within regions while using logistic chaotic sequences to trigger controlled global exploration and cross-region jumps. We instantiate CLDE for both single-objective and multi-objective multimodal optimization and evaluate it on standard benchmarks. Results demonstrate that CLDE consistently improves mode coverage, convergence stability, and structural diversity preservation across diverse multimodal landscapes. This paper makes the following contributions: • We propose CLDE, a region-centric framework that closes the loop over decision-space basins to bridge the decision–objective gap in multimodal optimization. • We develop persistence-guided region discovery to decode stable basins from a decision-space neighborhood graph and suppress spurious/shallow modes. • We introduce region saliency and adaptive budget allocation to focus evaluations on important basins while preserving minor-but-promising ones. • We design a region-aware variation with logistic-chaos perturbations for controlled cross-basin exploration and within-basin refinement. 2. Related Work The proposed CLDE framework is most closely related to three research lines: fitness landscape analysis and structure decoding, single-objective multimodal evolutionary algorithms, and multimodal multi-objective evolutionary algorithms. Although these directions have developed effective mechanisms for preserving multiple modes, it remains nontrivial to (i) decode stable decision-space regions online under limited evaluations and (ii) translate such structural signals into region-level resource allocation within an evolutionary loop. 2.1. Fitness landscape analysis and structure decoding Fitness landscape analysis (FLA) aims to characterize macroscopic properties of optimization problems, including ruggedness, correlation length, and the distribution of local optima, and is often used for diagnosis and algorithm comparison ( 45 ; 34 ; 29 ) . Classical FLA tools (e.g., random-walk correlation and information-theoretic indicators) provide global summaries of difficulty but typically do not reconstruct basin-level structures explicitly. To obtain a more structural view, local optima networks represent local optima as nodes and transitions between their basins as edges, revealing funnel structures and escape pathways on combinatorial landscapes ( 24 ) . For continuous problems, several studies estimate the number and size of basins by performing local search from random starting points and counting distinct local optima and their attraction domains ( 30 ; 16 ) . Graph-based approaches construct neighborhood graphs over samples and analyze connectivity or contour structures to infer the existence of multiple basins ( 33 ) . In parallel, topological data analysis (TDA) offers robust tools to extract features from scalar fields and point clouds. Persistent homology tracks the birth and death of connected components, loops, and voids during a filtration and separates short-lived noise from persistent structures ( 10 ; 3 ) . For scalar-valued fitness functions, 0 -dimensional persistent homology is closely related to the persistence of connected components in excursion sets, which naturally reflects the emergence and disappearance of peaks and basins. Nevertheless, much of the above structure analysis is used in an offline manner and often assumes relatively dense sampling; integrating such structural decoding tightly into an online evolutionary process under strict evaluation budgets remains less explored. 2.2. Single-objective multimodal evolutionary algorithms SOMMOP aims to locate multiple local and global optima of a scalar objective instead of a single best solution. Classical genetic niching methods include fitness sharing ( 12 ) , Mahfoud’s general niching framework ( 21 ) , clearing ( 28 ) , and speciation-based genetic algorithms ( 17 ) . These approaches introduce sharing functions or restricted competition so that crowded regions are penalized, enabling several peaks to be preserved simultaneously; comprehensive surveys summarize typical mechanisms and failure modes ( 16 ) . Differential Evolution (DE) is also a popular base for multimodal search due to its simple structure and self-adaptive steps. Crowding DE applies crowding selection so that each offspring competes against its nearest neighbor, promoting niche coexistence ( 35 ) . Neighborhood-based mutation