Goal edit

My PhD requires that I build a computer with the following attributes

  • models virtual neurons
  • neuronal interaction carried out by virtual synapses
  • virtual NGF ensures plasticity of neurons

Nature of Challenge: Putting Plasticity in a Silicon Cortex edit

As per our November 21st meeting, my supervisor has requested that I write an informal document to outline some of the challenges my PhD will eventually address. I am to present an outline December 4 and have it completed around December 15 The document must be 4000 words (no more, no less, and each word must be polysylabic), and must answer the below three questions:

  1. What is plasticity in Bioligy
    1. Neuroplasticity
    2. synaptic plasticity
    3. Spike timing dependent plasticity
    4. Long-term potentiation
    5. Long-term depression
    6. Metaplasticity
  2. What are the models around for plasticity
    1. General models?
    2. Network implementations
      1. Perceptron
      2. ADALINE
      3. Radial basis function
      4. Self-organizing map
      5. Simple recurrent network
      6. Hopfield net
      7. Boltzmann machine
      8. Committee of machines
      9. Associative Neural Network (ASNN)
  3. What are the challanges of running them tractably
    1. I am probably going to run through all of them and include running times.

Current Implementations edit

  1. Networks
    1. FeedForward Neural Network
      1. Single-layer perceptron
      2. Multi-layer perceptron
      3. ADALINE
      4. Radial basis function (RBF)
      5. Kohonen self-organizing network
    2. Recurrent network
      1. Simple recurrent network
      2. Hopfield network
    3. Stochastic neural networks
      1. Boltzmann machine
    4. Modular neural networks
      1. Committee of machines
      2. Associative Neural Network (ASNN)
  2. Nodes
    1. Binary Neuron