- chainer.functions.cross_covariance(y, z, reduce='half_squared_sum')¶
Computes the sum-squared cross-covariance penalty between
The output is a variable whose value depends on the value of the option
reduce. If it is
'no', it holds the covariant matrix that has as many rows (resp. columns) as the dimension of
y(resp.z). If it is
'half_squared_sum', it holds the half of the Frobenius norm (i.e. L2 norm of a matrix flattened to a vector) of the covarianct matrix.
A variable holding the cross covariance loss. If
'no', the output variable holds 2-dimensional array matrix of shape
N) is the number of columns of
z). If it is
'half_squared_sum', the output variable holds a scalar value.
- Return type
This cost can be used to disentangle variables. See https://arxiv.org/abs/1412.6583v3 for details.