### Refine by Time Frame

updated almost 2 years ago

### Guangdi Li (view profile)

For learning "Directed Maximum Spanning Tree", Chu-Liu/Edmonds Algorithm is implemented here.

DirectedMaximumSpanningTree( OriginalCostMatric,Root )

DirectedMinimalSpanningTree( OriginalCostMatric,Root )

MaximalDirectedMSF( CostMatric )

updated almost 2 years ago

### Guangdi Li (view profile)

This is the code of Cooper's K2 algorithm proposed in 1992, quick and convenient for using.

ConstructLGObj( OriginalSample )

GClosedFun( LGObj, X, PAX )

k2( LGObj,Order,u )

updated 3 years ago

### Guangdi Li (view profile)

Do you have problem with Fisher's exact test where the contingency table is more than 2*2?

FisherExactTest( XVector,YVector )

FisherExactTest_R( X,Y )

ControlCentor.m

updated 4 years ago

### Guangdi Li (view profile)

To discrete continuous data, CACC is a promising discretization scheme proposed in 2008

CACC_Discretization( OriginalData, C )

ControlCenter.m

updated 4 years ago

### Guangdi Li (view profile)

CAIM (class-attribute interdependence maximization) is designed to discretize continuous data.

CAIM_Discretization( OriginalData, C )

CAIM_Evaluation( OriginalData, C, Feature, DiscretInterva...

DiscretWithInterval( OriginalData,C,Column,DiscretInterva...

updated 5 years ago

### Guangdi Li (view profile)

Pearl's famous causal polytree recover algorithm is implemented here.

IsLeaf( Graph,Node )

Pearl_PolytreeRecover( CIMatric,CPMatricSkeleton )

ControlCenter.m

updated 5 years ago

### Guangdi Li (view profile)

For beginners, who loves Fisher's Exact Test but hates the theory behind.

[ Pvalue ConTable ]=FisherExactTest22( x,y )

ControlCenter.m

updated 5 years ago

### Guangdi Li (view profile)

We extend normal gray code to conside about different range for each digit.

GenerateGrayCode( Range )

GrayNumberConvertion(Value,Base,Digits )

ImprovedGenerateGrayCode( Range )

updated 5 years ago

### Guangdi Li (view profile)

Code for marginally and conditional mutual information in probability and information theory

ConditionalProbability( MDCObj,CurrentVar,VarParent,Vecto...

ConditionallyIndependent_MutualInformation( LGObj,Var1,Va...

ConstructLGObj( OriginalSample )

updated almost 6 years ago

### Guangdi Li (view profile)

NPC algorithm is designed for learning Bayesian network formed as DAG in 2001, by Steck

CITest_ChiTwoVar( MI, R, M, a )

ConditionallyIndependent_MutualInformation( LGObj,Var1,Va...

ConstructLGObj( OriginalSample )

updated almost 6 years ago

### Guangdi Li (view profile)

Chu-Liu-Edmonds Algorithm for learning "Undirected Maximum Weight Spanning tree" is proposed here.

UndirectedMaximumSpanningTree (CostMatrix)

ControlCentor.m

updated 6 years ago

### Guangdi Li (view profile)

It is credited as 1th algorithm applying stochastic sampling to belief updating in Bayesian networks

BinaryPLS_GenerateRandomData( GraphMatric,Number )

CompDecilToInt( RangeOfVar,CompDecNumber )

GenerateDAG( Node )