Home Machine Learning Ant Colony Optimization — Instinct, Code & Visualization | by James Koh, PhD | Jan, 2024

Ant Colony Optimization — Instinct, Code & Visualization | by James Koh, PhD | Jan, 2024

0
Ant Colony Optimization — Instinct, Code & Visualization | by James Koh, PhD | Jan, 2024

[ad_1]

The place it stands out from different swarm algorithms

This text is a continuation of my nature-inspired sequence.

Beforehand, I talked about Evolutionary Algorithm (EA), Particle Swarm Optimization (PSO), in addition to Synthetic Bee Colony (ABC). Nature is all over the place, and there’s actually extra areas the place people can profit by studying from nature.

At present, we concentrate on ants.

As kids, we learnt that ants are hardworking and cooperative. What our dad and mom hadn’t taught us was that ants collectively type a extremely refined swarm that communicates with each other successfully.

Data of ants or pheromones (or any diffusion of any chemical compounds) is just not required right here in any respect. These are simply names used for the aim of packaging. I’ve proven beforehand that you do not want the slightest information of a bee’s waggle dance with the intention to admire or make the most of ABC, nor do you should find out about genes or mutations or copy to use EA.

All you want is an understanding of English to have the instinct, together with very fundamental math and python programming expertise. Whereas I can be exhibiting some arithmetic for completeness, which incorporates Greek symbols, it’s actually only for the aim of completeness. It could be an important pity if these technical-sounding phrases or symbols cease you from studying these nice algorithms, so do your self a favor and skim on.

Earlier than going into any math or code, and even how the algorithm works at a excessive stage, it is smart to see the relevance. In any case, if it doesn’t assist to unravel an issue, why hassle within the first place?

The traditional instance which lecturers or proponents of Ant Colony Optimization (ACO) use is the double bridge experiment [1], which exhibits that this algorithm can be utilized to seek out the shortest path between two factors.

(Picture of ant from DALL·E 3, put collectively by writer utilizing PowerPoint.) Higher pathway is shorter, and therefore has the next density of ants, than backside pathway.

Furthermore, it’s sturdy to modifications within the surroundings. If present paths get obstructed, and/or if new paths come up, the answer may be up to date with ease, as a substitute of re-computing every thing from scratch.

[ad_2]