| |
Methods defined here:
- __init__(self, ni, no, f='linear')
- Set up instance of RBF net. N.B. RBF centers and variance are selected at training time
Input:
ni - <int> # of inputs
no - <int> # of outputs
f - <str> output activation fxn
- err_fxn(self, w, X, Y)
- Return vector of squared-errors for the leastsq optimizer
- fwd_all(self, X, w=None)
- Propagate values forward through the net.
Inputs:
inputs - vector of input values
w - packed array of weights
Returns:
array of outputs for all input patterns
- pack(self)
- Compile weight matrices w,b from net into a
single vector, suitable for optimization routines.
- test_all(self, X, Y)
- Test network on an array (size>1) of patterns
Input:
x - array of input data
t - array of targets
Returns:
sum-squared-error over all data
- train(self, X, Y)
- Train RBF network:
(i) select fixed centers randomly from input data (10%)
(ii) set fixed variance from max dist between centers
(iii) learn output weights using scipy's leastsq optimizer
- unpack(self)
- Decompose 1-d vector of weights w into appropriate weight
matrices (self.{w/b}) and reinsert them into net
|