Biofunctionality: a novel learning method for intelligent agents

Growing a knowledge base for an intelligent agent is the main concern in developing a learning strategy. While simple in structure, rigid and mathematically precise learning models are generally ineffective in expressing complex operating environments. A learning model envisioned for use in the physical world, should also be reasonably easy to implement. Our daily lives and experiences suggest that a human-like learning strategy with all its flexibilities is better suited for successful functioning in a hard-to-model environment. A rule-based learning model, which follows the learning patterns of the humans, contains the characteristics mentioned above. Here, we are presenting the biofunctional learning model and its implementation using the classifier systems.