Two meanings of “many”

Multiple objectives and multiple modes solve different decision problems. A serious architecture must preserve the distinction.

MULTI-OBJECTIVE

Many definitions of quality

A candidate is assessed on several criteria—cost, time, quality, risk, energy or fairness. The desired output is a set of non-dominated trade-offs.

one search space → many objective values
MULTIMODAL OPTIMISATION

Many structurally different solutions

Several permutations may have equal or near-equal objective value while representing different routes, schedules, assignments or operational strategies.

one quality region → many solution structures
Terminology: here, multimodal refers to multiple optima or solution modes. It does not mean multimodal machine-learning input such as image, text and audio.

What multi-objective COIN already changes

Single-objective COIN can rank candidates by one scalar fitness value. With several objectives, no single scalar ordering is neutral. Improving one criterion may worsen another. The dissertation therefore combines the COIN generator with non-dominated sorting and crowding distance inspired by NSGA-II.

quality ↑
cost →

Illustrative objective space. The highlighted candidates form a non-dominated trade-off boundary; none is universally best without a preference.

Non-dominated rank defines quality

Candidate A dominates B when A is no worse in every objective and better in at least one. The first front contains candidates not dominated by any other candidate. Subsequent fronts describe progressively weaker trade-off layers.

Crowding distance protects spread

Rank alone can concentrate selection around a small region of the Pareto front. Crowding distance favours candidates in less populated objective-space neighborhoods, helping the evidence set cover more trade-offs.

COIN then learns from the selected trade-offs

The better and worse groups become multi-objective evidence interfaces. Incidences in well-ranked, diversity-supporting candidates are rewarded. Incidences in weak candidates may be punished. The matrix is still an edge model, but its evidence now arrives from a set of compromises rather than a single ordering.

Multi-objective learning architecture

01Sample from HGenerate feasible permutations
02Evaluate vectorf(x) = [f₁, f₂, …, fₖ]
03Sort frontsNon-dominated rank
04Protect spreadCrowding distance
05Form evidenceBetter / worse groups
06 · LEARNReward and punishment update the shared probability matrix HThe next population reflects accumulated evidence from several trade-offs.

The shared matrix is not merely an average of elite solutions. Its negative update changes the geometry of the search: coincidences repeatedly shared by poor candidates lose probability, so the generator spends less effort reconstructing the same unproductive substructures. At the same time, distinct positive building blocks belonging to different high-quality modes can retain probability in H.

Central systems idea

COIN supports diversity not by assigning one generator to each mode, but by pruning shared negative structure while allowing several useful structures to coexist in one generator.

How a shared COIN generator retains multimodality

Suppose two production schedules have similar cost and makespan but use different worker allocations. Or two routes have nearly identical distance but use different roads. These are distinct modes: operationally meaningful alternatives that may occupy similar positions in objective space.

The important mechanism is selective suppression. If poor candidates from several regions repeatedly contain the same negative coincidence, punishment reduces that shared negative building block in H. This prunes a common route into weak regions of the search space. The building blocks that distinguish Mode A from Mode B do not need to be erased merely because they differ; when they remain associated with strong candidates, both can retain probability.

MODE A RETAINED1 → 3 → 4 → 6Positive building blocks remain available
SHARED Hnegative-space pruningShared poor incidences are suppressed across the search
MODE B RETAINED1 → 5 → 2 → 6A different useful structure can coexist

Multimodality is therefore difficult, but it does not require splitting the original COIN generator. It depends on whether reward preserves the structures of several strong modes and punishment removes negative structures without indiscriminately destroying their alternatives.

How negative-space pruning supports several modes

COIN keeps one shared generator. Its multimodal capability depends on what the update removes and what it leaves available. The useful unit of reasoning is not a complete solution but the coincidence building blocks distributed across the population.

  1. Observe diverse strong candidates.

    Different high-quality modes contribute different useful coincidences to the same H.

  2. Find recurring negative structure.

    Poor candidates may differ globally while sharing local incidences that repeatedly lead toward weak regions.

  3. Punish shared negative building blocks.

    Probability is removed from those incidences and returned to their alternatives, pruning common paths into low-quality space.

  4. Retain modal building blocks.

    Useful coincidences belonging to different modes can remain probable when the evidence does not identify them as negative.

  5. Resample combinations.

    The shared H continues generating candidates from the reduced search space while preserving several promising structural vocabularies.

  6. Repeat under population evidence.

    Multimodality emerges from accumulated selective pressure rather than an explicit one-generator-per-mode partition.

Diverse evaluated populationseveral strong modes may coexist with recurring weak structures
reward useful · punish shared negative
MODE AA+Useful building blocks retained
SHARED HHCommon negative building blocks suppressed
MODE BB+Different useful building blocks retained
constrained sampling from one model
A pruned search space that can still express diverse high-quality solutions

The generator remains shared

The model is not split by mode. What changes is the probability landscape: negative correlations reduce access to repeatedly weak substructures, while the surviving positive structures can describe more than one basin of strong solutions.

The capability is empirical, not automatic

A shared H can retain several modes only when the representation and update distinguish their useful building blocks from the structures associated with poor candidates. Learning rate, population composition and better/worse selection therefore affect whether diversity survives.

How to evaluate both capabilities

QuestionMulti-objective evidenceMultimodal evidence
QualityHypervolume, epsilon indicator, generational distancePeak ratio, success per mode, best quality within each niche
DiversitySpread and coverage in objective spaceEdge-set diversity, permutation distance, number of stable modes
ReliabilityFront stability across seeds and budgetsMode survival, rediscovery rate and collapse frequency
EfficiencyEvaluations and time to a useful frontEvaluations required to discover and retain each mode
Decision valueCoverage of stakeholder trade-offsOperationally distinct alternatives after removing duplicates

A strong experiment should compare positive-only learning with shared-generator COIN using reward and punishment. The central ablation is whether negative-space pruning in one H retains more modes, retains them longer and reaches them more reliably across random seeds.

Why decision-makers may need both

PRODUCTION

Schedule portfolios

Balance makespan, labour, energy and changeover while preserving different workforce or machine-allocation strategies.

LOGISTICS

Route alternatives

Trade distance, time, emissions and risk while preserving geographically distinct routes resilient to disruption.

HEALTH OPERATIONS

Resource configurations

Balance access, waiting time and workload while retaining alternatives suitable for different demand scenarios.

ENGINEERING DESIGN

Structurally distinct designs

Offer several high-quality configurations rather than minor variants of the same architecture.

The practical output is not “the algorithm found many answers.” It is a compact portfolio where each alternative represents a meaningful trade-off or operational mode. Redundant solutions should be removed; differences should be explained in language stakeholders can evaluate.

Capability boundary and research status

COIN uses a shared generator: its multimodal capability is connected to negative knowledge suppressing shared negative building blocks while useful building blocks of diverse solutions remain represented.

The architecture retains a representation boundary. H is a first-order pairwise model. Problems governed by higher-order blocks, long-range precedence or context-dependent interactions may require richer probabilistic structures or hybrid local search.

← Detailed architectureNext: Empirical Evidence →