fcMeanVar: Mean and covariance vectors of units (many observations)

In mgoulet847/tagi: Tractable Approximate Gaussian Inference in Neural Networks

Description

This function calculate the mean vector of units μ_{Z} and the covariance matrix of the units Σ_{Z}for a given layer.

Usage

fcMeanVar(mz, Sz, mw, Sw, mb, Sb, ma, Sa, ni, no, B, rB)

Arguments

mw	Mean vector of weights for the current layer
Sw	Covariance of weights for the current layer
mb	Mean vector of biases for the current layer
Sb	Covariance of biases for the current layer
ma	Mean vector of activation units from previous layer
Sa	Covariance of activation units from previous layer
ni	Number of units in previous layer
no	Number of units in current layer
B	Batch size
rB	Number of times batch size is repeated

Value

• Mean vector of units for the current layer μ_{Z}

• Covariance matrix of units for the current layer Σ_{Z}

mgoulet847/tagi documentation built on Dec. 21, 2021, 5:10 p.m.