Refresh weights of neural network without `Configure` function in MATLAB

As you know we can use configure function in MATLAB to initial weights and biases but this function configure all networks (weights/biases + other properties). I want after configuring network using configure, have a loop and initial weights for 10 times and store these weights in a cell array to use it in another part of network. (Initial weights and biases based on initial function of network object)

Is any function to only refresh initial weights and biases without configuring it again and again to accelerate the speed?

Thanks.

Answer 1

[Greg Heath]

Use the init function

 init(net);

Hope this helps

Thank you for formally accepting my answer

Greg