The SUMO toolbox  2018a
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Pages
Public Member Functions | Public Attributes | List of all members
LOLASampleRanker Class Reference

A class that rates samples according to the LOLA non-linearity criterion. More...

Inheritance diagram for LOLASampleRanker:
Inheritance graph
[legend]

Public Member Functions

function LOLASampleRanker (var varargin)
 
function setDimensions (var inDim, var outDim)
 Sets the dimensions. More...
 
function calculateError (var samples, var values)
 Calculate error is overridden to produce the true gradient error. More...
 
function getAdditionalCandidates (var samples, var A)
 Perform some kind of heuristic to generate additional points near point A so that a better candidate can be selected. More...
 
function getNeighbourhoods ()
 Getter for the neighbourhoods of the points. More...
 
function getDesignType ()
 
function selectSamples (var state)
 

Public Attributes

var inDim
 
var outDim
 
var neighbourhoodSize
 
var gradientMethod
 
var fastNeighbourhoodCalculation
 
var debug
 
var combineOutputs
 
var sampleSize
 
var neighbourhoods
 
var neighbourhoodScores
 
var neighbourhoodMaxDistance
 
var gradients
 
var test
 
var gradientErrors
 
var neighbourhoodErrors
 
var averageGradientError
 
var neighbourhoodSubLeftSide
 
var neighbourhoodSubRightSide
 
var neighbourhoodSubIndexArray
 
var fastDistance
 
var inputWeights
 
var nScores
 
var nSkippedScores
 
var isClassification
 
var numberOfClasses
 

Detailed Description

A class that rates samples according to the LOLA non-linearity criterion.

Constructor & Destructor Documentation

function LOLASampleRanker ( var  varargin)
inline

Member Function Documentation

function calculateError ( var  samples,
var  values 
)

Calculate error is overridden to produce the true gradient error.

function getAdditionalCandidates ( var  state,
var  A 
)

Perform some kind of heuristic to generate additional points near point A so that a better candidate can be selected.

function getDesignType ( )
inline
function getNeighbourhoods ( )

Getter for the neighbourhoods of the points.

function selectSamples ( var  state)
inline
function setDimensions ( var  inDim,
var  outDim 
)
inline

Sets the dimensions.

The sampleselector processes this once and accounts for frequency variables

Member Data Documentation

var averageGradientError
var combineOutputs
var debug
var fastDistance
var fastNeighbourhoodCalculation
var gradientErrors
var gradientMethod
var gradients
var inDim
var inputWeights
var isClassification
var neighbourhoodErrors
var neighbourhoodMaxDistance
var neighbourhoods
var neighbourhoodScores
var neighbourhoodSize
var neighbourhoodSubIndexArray
var neighbourhoodSubLeftSide
var neighbourhoodSubRightSide
var nScores
var nSkippedScores
var numberOfClasses
var outDim
var sampleSize
var test

The documentation for this class was generated from the following files: