Hierarchical Learning

The ant colonies show specialization and oriented-collective behavior to solve problems. The individuals in this comparison interact with a single domain when cutting leaves, transporting them to the house, and storing them. Each individual has a specialized function: cutter, transporter, farmer, etc., in the same domain. The preliminary learning algorithm is a collaborative learning approach. When an ant seeks a leaf, it exchanges information with each other to share the local from harvest home. Each individual is trusted due to the pheromone. Therefore, the pheromone acts as exponential feedback to guarantee the most current shortest path between the ant and the leaf place. On the other hand, the competition between individuals in the same function (leaf cut) contrasts with the cooperation that promotes their evolution.

In this manner, the individual shares the surrounding knowledge with the collective. Also, the pheromone has a joint identity function. The soldiers have a colony protection responsibility if any individual detects an intruder due to this identity violation. The soldiers and all individuals in the anthill are trying to kill the invader. This behavior is a collaborative action where different non-specialized individuals execute an inter-domain function, i.e., a function not specialized with a trying-error approach to address a collective problem [1].