QLinear

Full name Microsoft.VisualBasic.MachineLearning.ComponentModel.Activations.QLinear Assembly Microsoft.VisualBasic.MachineLearning Members 4

A quadratic-linear activation function: f(x) = ln(x ^ 2) when x >= 1, otherwise zero.

00 Remarks

The output of this function is flat (zero) on the interval (-inf, 1) and grows logarithmically outside of that interval.

01 Syntax

Microsoft.VisualBasic.MachineLearning.ComponentModel.Activations.QLinear

02 Methods

Name	Overloads	Summary
Function	1	Calculates the function value: ln(x ^ 2) when x >= 1, otherwise 0.
ToString	1	Display this activation function as a text expression.
Derivative	1	Calculates the derivative: f'(x) = 1 / (2 * x).

03 Properties

Name	Overloads	Summary
Store	1	Gets the XML serializable data model of this quadratic-linear function.

04 Members

method Function #

Function(Double)

Calculates the function value: ln(x ^ 2) when x >= 1, otherwise 0.

Parameters

Name	Type	Description
`x`	`Double`	The function input value.

Returns

The function output value.

method ToString #

ToString

Display this activation function as a text expression.

Returns

The text expression of this quadratic-linear function.

method Derivative #

Derivative(Double)

Calculates the derivative: f'(x) = 1 / (2 * x).

Parameters

Name	Type	Description
`x`	`Double`	The function input value.

Returns

The derivative value; a large constant value 10000 will be returned when x is zero, so that the derivative value will never be infinite.

property Store #

Store

Gets the XML serializable data model of this quadratic-linear function.

Returns

A ActiveFunction data model which its function name is QLinear and no argument is required.