Conclusion
- Spiking nets are different from perceptrons largely because
of time.
- There is some sound evidence that the brain, using spiking
neurons, does some learning with STDP.
- The old mechanism used for the students system works, but
has biologically implausible presentation.
- The competitive net mechanism is more biologically plausible.
- The results for the student system are reasonable, and the
competitive net is above chance.
- We are working on furthering our understanding of competitive
nets.
- The hope is that we can learn more about how the brain works,
and build better machine learning systems with this knowledge.
