aggregations

Aggregations of arrays (vectors, tensors) into lower dimensions.

Functions

aggregate(mean_loss[, alias])	General aggregation of across multiple output dimensions given an alias string or an input callable.
convolution_aggregate(x[, epsilon])	Generalized convolutional aggregation over flattened features.
geometric_mean(x)	Computes the geometric mean over all elements of the input array.
log_mean(x)	Computes the mean of the log of all elements of the input array.
weighted_sum(x[, weights])	Reduces the value over the sample (batch) dimension, applying weights if provided.