Ant colony optimization
Algorithms AIMech:AntColonyOptimization DRAFT
Aliases: ACO, Ant colony optimization metaheuristic
A metaheuristic framework for optimization in which artificial ants construct solutions using local information and communicate indirectly through shared pheromone information that is adapted from their search experience.
Identity Notes & Concept Scope
The record denotes the ACO framework defined in the 1999 source, not a single fixed update equation. Ant System, ACS, MAX–MIN and AntNet are individual procedures with additional choices. The source distinguishes strict ACO instantiations from the broader class of ant-inspired algorithms, so all ant algorithms are not treated as synonyms. PBIL’s probability-vector update and other population heuristics share some ideas without becoming identical to ACO. No normalized identity match was found across all 442 pinned AIO labels/scoped synonyms and 295 preceding native IDs/names/aliases. Full-checkout fixed-string searches included ignored and hidden files in worktree and main, returning 57 and 57 lines respectively. Text hits were prior candidate leads and neighboring-method mentions in maintained records, saved reviews and discovery reports; they did not identify an existing record for these candidates. Six exact AIO CLASS nodes with scoped synonyms/comments and outgoing edges and six native neighboring mechanisms were inspected. Absence is bounded to this snapshot and semantic neighborhood.
Claims & Grounded Evidence
1 Claim
Mechanism
#method
Represent the problem through states or solution components, feasible local moves and an objective. Artificial ants incrementally construct a solution or a component of one, maintaining private history as needed for feasibility. Their stochastic decisions use locally available pheromone information and, where configured, problem-specific heuristic information. Search experience changes the shared pheromone representation so later constructions can exploit prior discoveries. The framework schedules ant generation/activity, pheromone evaporation or forgetting, and optional globally informed daemon actions such as extra pheromone updates or local optimization. Deposits can occur during construction, after a solution is evaluated, or through the daemon; the ant’s stepwise and delayed deposit routines can both be absent when the daemon supplies deposits. A concrete algorithm must specify the representation, decision policy, update rule, scheduling and stopping criterion. Sequential or concurrent schedules are permitted by the prose; the framework does not prescribe a universal tau-power formula, inverse-cost deposit, elite rule or trail bound.
⚠️ Curated Limitations & Negative Results
- A biological analogy is not a guarantee of optimality, exploration or prevention of stagnation. Positive feedback can concentrate search, and stochastic transitions or evaporation alone do not ensure that every feasible solution remains reachable. Constraints, candidate restrictions, zero weights and finite precision matter.
- Heuristic information and daemon actions are optional or instance dependent. Source prose permits local optimization, lookahead and other extensions; these are not obligatory baseline operations. Its Figure 3 caption refers to an in_parallel statement while the displayed code uses schedule_activities, so the scheduling prose and explicit instantiation must govern an executable implementation.
- The framework was synthesized from algorithms with varied representations and update schedules. Membership does not transfer one member’s convergence result or benchmark performance to another. The static/dynamic application examples do not define a universal network-routing or machine-learning algorithm.
- The paper’s broad PBIL/evolutionary/learning-automata comparisons are conceptual parallels, not evidence of exact algorithm equivalence. No experimental outcomes, fixed best parameterization or full field-wide coverage claim is curated.
Artificial Intelligence Ontology (AIO) Grounding
Explicit AIO Snapshot Assessment: No Match
Rationale: No suitable AIO target was found for the ACO framework in the inspected snapshot and semantic neighborhood. Instance-specific learning relations and model labels are not generalized into a framework identity.
Search Scope: All 442 pinned AIO labels/scoped synonyms and 295 preceding native identities, ignored/hidden full-checkout text, six exact AIO CLASS targets and six neighboring native methods.
| Examined AIO Candidate | Rejection Reason |
|---|---|
AIO:TrainingStrategy
|
The framework is general constructive optimization, not specifically preprocessing or training a machine-learning model. |
AIO:MachineLearningTask
|
The target is a learning task, whereas this record defines a general search framework; optimization terminology alone does not identify that task. |
AIO:ReinforcementLearning
|
Some ACO members have explicit reinforcement-learning formulations, but the framework does not require their action-value or policy-learning semantics. Their individual related matches do not establish an appropriate relation for every ACO instantiation. |
AIO:ReinforcementLearningLLM
|
Artificial ants and shared search memory do not establish a reward-fine-tuned language-model identity. |
AIO:EvolutionaryLLM
|
The target is a large language model; population-search analogies and its related evolutionary-algorithms synonym do not identify this framework. |
AIO:MarkovChain
|
A stochastic augmented-state representation does not make this constructive search framework the pinned network/model class. Its representation and transition laws remain instance dependent. |
Evidence Sources & Primary Literature
1 Source| Source ID | Publication / Document | Type / Access | Reference & Checksum |
|---|---|---|---|
method-source
|
Marco Dorigo, Gianni Di Caro, Luca M. Gambardella
|
primary paper
full_text
|
ID: doi:10.1162/106454699568728
sha256:7c6a5a0df5dd…
|
Curation Provenance
Curator: Codex
Method: ai_assisted
Model: GPT-6
Created: 2026-10-11T13:47:28Z
Updated: 2026-10-11T13:47:28Z
Source File:
data/curated/AntColonyOptimization.yaml