zxzhijia / Brian2STDPMNIST

Brian 2 version of Paper "Unsupervised Learning of digit recognition using STDP"
129 stars 72 forks source link

Why the learned weights can be decreased gradually #7

Closed hustyanghui closed 6 years ago

hustyanghui commented 6 years ago

Hello ,everyone , I run the simulation (I have not change any parameter).However,when the neuron learned a new number ,the learned weights will be decreased.there is no such setting in "STDP".As the file shows .Firstly,we can see the neuron learned a number "5",soon,the neuron learned number "0" again,but the learned "5" was decreased ,and the background of the input picture was decreased to zero.How did it work? Is there anyone meet the problem?Would you like to give me some solutions?Thank you!

1 2