Skip to content

隐含层权值调整公式好像错了? #2

@attackon1point6

Description

@attackon1point6

class Layer下的get_error函数,我认为_error内容应该是这样的:
neru0=self.units[j]
sum=0
for i in range(len(neru0.weight)):
sum+=neru0.weight[i]*deltas[i]
return sum
因为隐含层权值调整应该是同一个隐含层神经元对所有输出层神经元误差的求和,不是一个输出层神经元对所有隐含层的求和

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions