Tune in to the first Women in ML Symposium this Tuesday, October 19 at 9am PST Register now


View source on GitHub

Adds a maxout op from https://arxiv.org/abs/1302.4389

"Maxout Networks" Ian J. Goodfellow, David Warde-Farley, Mehdi Mirza, Aaron Courville, Yoshua Bengio

Usually the operation is performed in the filter/channel dimension. This can also be used after fully-connected layers to reduce number of features.

inputs Tensor input
num_units Specifies how many features will remain after maxout in the axis dimension (usually channel). This must be a factor of number of features.
axis The dimension where max pooling will be performed. Default is the last dimension.
scope Optional scope for variable_scope.

A Tensor representing the results of the pooling operation.

ValueError if num_units is not multiple of number of features.