Swarm Fuzzy-Reinforcement Coordination using Bloom's Taxonomy of the Cognitive Domain

From Bloom's taxonomy, cognitive aspect of learning is defined in six phases of knowledge, comprehension, application, analysis, synthesis, and evaluation. Imitating this learning structure, we propose here a Swarm Fuzzy-Reinforcement Coordination (SFRC) strategy for targeted delivery of cancer drug. The swarm of learning nanomachines delivers the drug to the cancer site in a coordinated manner based on the Bloom's cognitive phases. Using these six phases, we define different cognitive single and social states for each of the nanomachines. These states are defined by applying a fuzzy inference system (FIS). At each cognitive level, each nanomachine is allowed to socially join the swarm and start coordinating, communicating, and doing collective therapy. Using the distance toward the cancer site, the quality of each swarm is classified. Those swarms that are closest to the cancer site begin to release oxygen into the environment in order to communicate the vicinity of cancer site to other nanomachines in other swarms. Nanomachines in the wrong swarm disjoin from the group and join the swarm with the better location toward the cancer site. Here we use the Vascular Endothelial Growth Factor (VEGF) concentration, which is high around the cancer site, as a controlling signal. Aiming to decrease VEGF, swarms of nanomachines coordinate with each other using a reinforcement learning concept. Result comparison with mathematical therapy model shows the merits of the SFRC method.