You must login before you can run this tool.
[Illinois]: Posterior target probability given single-sensory input (delta rule)
Train a single sigmoidal unit using the delta rule to estimate posterior target probability given sensory input of one modality (i.e., visual)
Figure 9.7 depicts a single, sigmoidal unit used to simulate a unisensory neuron in the superior colliculus. The unit y receives input x and bias b. The weight matrix V in this case is a row vector containing the weights of the two connections to y, one from x and the other from b. The unit computes the weighted sum of its input and passes the result through the sigmoidal squashing function (see Chapter 6). The input x could represent the visual input to the collicular neuron, and bias b could represent a constant influence on neural firing rate due to the biophysical properties of the neuron or to nonspecific inputs that are constant in combination.
Researchers should cite this work as follows: