Conclusion
- We use a reasonably biologically faithful neural model, learning
rule, and topology.
- We present continuous value input.
- There are neuropsychological flaws to the model.
- We categorise nearly as well as standard machine learning
algorithms.
- Spontaneous firing in unstimulated regions emerge from
hypo-fatigue.
- The pre-compensatory and post-compensatory learning rules
along with the spontaneous firing spread neural circuits
into unused areas.
