Artificial Bee Colony (ABC)
ABC models honey-bee foraging, first described by Dervis Karaboga (2005). Three roles divide the explore/exploit labor: employed bees exploit known food sources, onlooker bees pick sources by advertised quality (the "dance"), and scout bees abandon exhausted sources to explore new areas. Source: andrey-dik's MQL5 implementation (plus a modified "swarm" variant using fixed-size bee clusters per area rather than dynamic allocation).
How it works
Employed bees search near current food sources; onlookers probabilistically concentrate on the higher-fitness ones; when a source stops improving it is abandoned and a scout reseeds randomly — an explicit abandonment-driven exploration-vs-exploitation balance.
Canonical mechanics (Karaboga 2005)
The independent reference wikipedia-artificial-bee-colony pins down the original definition beyond
Dik's implementation. A new candidate position is generated by
vᵢₖ = xᵢₖ + Φᵢₖ·(xᵢₖ − xⱼₖ), with Φ random in [−1, 1] and j another randomly chosen source — a
neighbour-difference perturbation reminiscent of differential-evolution's difference vector. Onlookers
pick sources by roulette-wheel, fitness-proportional probability Pᵢ = fitᵢ / Σⱼ fitⱼ, and "the first
half of the swarm consists of employed bees, and the second half constitutes the onlooker bees." The one
control knob driving exploration is the limit: "if a position cannot be improved over a predefined
number (called limit) of cycles, then the food source is abandoned" and a scout reseeds — so ABC's
explore/exploit balance reduces to a single stagnation counter.
Benchmark (context-relative)
In this article's early scoring, ABC ranked 4th, scoring 0.49447 (classic) / 0.50836 (modified). Its signature is "phenomenal convergence for smooth and discrete functions with few variables" — 100% convergence on all five 2-variable tests — but "average scalability," decaying on 40- and 1000-param functions, with high parameter sensitivity and local-optima entrapment. Earlier, small-field methodology, not the mature %-of-MAX rating (see population-optimization-benchmark; no-free-lunch-theorem).
Why T2 (re-tiered 2026-08-25)
Held at T3 since ingest, which graded the venue — MQL5, a broker's community article site — rather than what the page rests on. This page's load-bearing claims are a mechanism description of an algorithm published by Karaboga (2005) and a comparative score, both from andrey-dik, who did not write the algorithm and has no stake in where it lands. That is somebody measuring somebody else's work, which QUALITY.md's ladder puts at T2.
The spoke's standing finding — that in this field the scorer is usually the author — is real and does not apply here; it applies to DOAm, which stays T3 because there Dik is scoring his own modification. The confound that does remain is implementation: Dik re-implements each optimizer in MQL5, so his figure grades his code as much as the method, and his scoring scheme changed mid-series (population-optimization-benchmark). Both stay flagged in prose, where a methodological limit belongs — not in the tier.
Note the floor is unmoved by this: T1 ÷ tiered counts T1 against the same denominator either way, so
the change alters the tier mix and not the spoke's floor of 0.18.
Related
metaheuristic-optimization · exploration-vs-exploitation · population-optimization-benchmark · no-free-lunch-theorem · andrey-dik · grey-wolf-optimizer · ant-colony-optimization · differential-evolution · wikipedia-artificial-bee-colony