Sensory Substitution & Plasticity

Neat article in today’s NYT on how a tongue stimulator can re-route many different sensory modalities. The most dramatic example (which the article focuses on) is how the device helps rehabilitate a patient with extensive vestibular damage and trains her brain to use whatever vestibular neurons she has left by this alternative (somatosensory) pathway. Click below for the full article.
Continue reading

Advertisements

Neural networks at your fingertips

This website has short, free, open-source C implementations of 8 kinds of neural networks:

  • Adaline Network
  • Backpropagation Network
  • Hopfield Model
  • Bidirectional Associative Memory
  • Boltzmann Machine
  • Counterpropagation Network
  • Self-Organizing Map
  • Adaptive Resonance Theory

    doing 8 kinds of tasks:

  • Classification of Digits 0-9
  • Prediction of the Annual Number of Sunspots
  • Associative Recall of Images
  • Association of Names and Phone Numbers
  • Traveling Salesman Problem
  • Determination of the Angle of Rotation
  • Pole Balancing Problem
  • Stability-Plasticity Demonstration
  • The symbol grounding problem and recurrent neural networks with parametric biases

    The symbol grounding problem (apparently) is: assuming that someone’s “cognitive” levels of mind work in terms of symbols, how to design the interface of the symbolic levels with the low-level sensory-motor systems?

    Jun Tani suggests a “recurrent neural network with parametric biases” (RNNPB).

    I haven’t had time to read further yet, but it looks very interesting so I’m passing it along now lest it get lost. When I get around to it I’ll post an update that summarizes what RNNPBs are and precisely how they interface symbol computation with lower-level systems. I may attend the talk (which is tomorrow).

    Read on for an abstract from Jun Tani’s talk.
    Continue reading