Function-Oriented Adaptive Pulse-Coded Network Using Model Reference Principle.
Code as a research output can either be uploaded directly from your computer or through the code management system GitHub. Versioning of code repositories is supported.
How can one design an adaptive pulse-coded neural network (PCNN) that is based on psycho-phenomenological foundations?
Many PCNNs (in neuroscience) are developed using scientific reduction. The practice of migrating scientific study and theory from the level of phenomena more directly observable by our senses to levels of increasingly refined scientific constructs. This is commonly referred to as, understanding things at the level of some postulated "absolute essence."
However, there are still holes in our understanding at the current state of research. For instance, specific mechanisms for potentiation in cell assemblies is still unclear. For the synthesis of knowledge of a system one must also consider model order reduction.
Using model reference principle (MRP), a well-known tool in control systems design, an adaptive PCNN is designed. This is done by incorporating phenomenological larger Scale, level-coded model adaptive properties. The reference model is Grossberg dipole network1. And the PCNN2 being adapted uses Eckhorn neural unit3.
- Grossberg S. A neural theory of punishment and avoidance, II: Quantitative theory. Mathematical Biosciences 1972; 15:253285. DOI: 10.1016/0025-5564(72)90038-7.
- Sharma B. Lungsi. Construction of a pulse-coupled dipole network capable of fear-like and relief-like responses. Connection Science 2016; 28:295-310. DOI: 10.1080/09540091.2016.1185393.
- Eckhorn R, Reitboeck HJ, Arndt M, et al. Feature linking via synchronization among distributed assemblies: Simulations of results from cat visual cortex. Neural Computation 1990; 2:293-307. DOI: 10.1162/neco.19126.96.36.1993.
Please read the Instructions_to_run_the_codes on how to run the codes.