Movatterモバイル変換


[0]ホーム

URL:


FRESA.CAD

Feature Selection Algorithms for Computer Aided Diagnosis.

Set of functions for: Conditioning, Feature Selection, MachineLearning, Cross-Validation, and Visual Evaluation

Table of Contents

Overview

The design of diagnostic or prognostic multivariate models via theselection of significantly discriminant features is complex.

FRESA.CAD provides a series of functions for: Data conditioning,Feature Selection, Machine Learning, Benchmarking, Visualization andReporting.

CategoryFunction(s)Purpose
Conditioning/PreprocessingnearestNeighborImpute()Impute missing values
Conditioning/PreprocessingFRESA.Scale()Data Scale/Normalization
Conditioning/PreprocessingfeatureAdjustment()Adjust variables removing collinearity
Conditioning/PreprocessingIDeA()/ILAA()Multicollinearity Mitigation
Feature SelectionuniRankVar()Univariate Analysis
Feature SelectionBSWiMS.model()Linear Model Subset Selection
Feature Selectionunivariate_BinEnsemble()Ensemble Select Top Features
Feature Selectionunivariate…Filter Select Top Features …
Machine LearningBSWiMS.model()Bootstrap Modeling
Machine LearningfilteredFit()Pipeline ML: Scale/Filter/Transform/Learn
Machine LearningHLCM()/HLCM_EM()Latent-Class Based Modeling
Machine LearningGMVECluster()Unsupervised Clustering via GMVE
Benchmarking / EvaluationRandomCV()Random Holdout Validation
Benchmarking / EvaluationBinaryBenchmark()Binary Model Evaluation
Benchmarking / EvaluationOrdinalBenchmark()Ordinal Model Evaluation
Benchmarking / EvaluationCoxBenchmark()Cox-based Model Evaluation
Visualization / ReportingRRPlot()Survival Model Evaluation
Visualization / ReportingpredictionStats_binary()Report Cross Validation Results Binary
Visualization / ReportingpredictionStats_Ordinal()Report Cross Validation Results Ordinal
Visualization / ReportingpredictionStats_survival()Report Cross Validation Results Survival

Besides the above listed functions the library provides predictorsand wrappers of common machine learning methods, and many otherauxiliary functions.

Installation

You can install the official release of the package from CRANusing:

install.packages("FRESA.CAD")

To install the development version from GitHub, use:

# Install 'devtools' package if you haven't alreadyinstall.packages("devtools")# Install the package from GitHubdevtools::install_github("https://github.com/joseTamezPena/FRESA.CAD")

Usage

#Load the packagelibrary(FRESA.CAD)#For comprehensive evaluaiton of confusion tableslibrary("epiR")# Example usagedata(stagec,package ="rpart")options(na.action ='na.pass')dataCancer<-cbind(pgstat = stagec$pgstat,pgtime = stagec$pgtime,as.data.frame(model.matrix(Surv(pgtime,pgstat)~ .,stagec))[-1])#Impute missing valuesdataCancerImputed<-nearestNeighborImpute(dataCancer)data(cancerVarNames)UniRankFeaturesRaw<-univariateRankVariables(variableList = cancerVarNames,formula ="pgstat ~ 1+pgtime",Outcome ="pgstat",data = dataCancer,categorizationType ="Raw",type ="LOGIT",rankingTest ="zIDI",description ="Description",uniType="Binary")print(UniRankFeaturesRaw)# A simple BSIWMS ModelBSWiMSModel<-BSWiMS.model(formula =Surv(pgtime, pgstat)~1, dataCancerImputed)#The list of all models of the bootstrap forward selectionprint(BSWiMSModel$forward.selection.list)#With FRESA.CAD we can do a leave-one-out using the list of modelspm<-ensemblePredict(BSWiMSModel$forward.selection.list,                          dataCancer,predictType ="linear",type="LOGIT",Outcome="pgstat")#Ploting the ROC with 95pm<-plotModels.ROC(cbind(dataCancer$pgstat,                               pm$ensemblePredict),main=("LOO Forward Selection Median Predict"))#The plotModels.ROC provides the diagnosis confusion matrix.summary(epi.tests(pm$predictionTable))

More examples of FRESA.CAD usage can be found at:https://rpubs.com/J_Tamez

Contributing

Contributions are welcome! If you’d like to contribute to thisproject, please follow these guidelines:

- Fork the repository.

- Create a new branch:git checkout -b feature/new-feature.

- Make your changes and commit them:git commit -m 'Add new feature'.

- Push to the branch:git push origin feature/new-feature.

- Submit a pull request.

License

This project is licensed under the LGPL>=2.0.

Contact

For any questions or feedback, feel free to contact us at:

Email: jose.tamezpena@tec.mx

Twitter:@tamezpena


[8]ページ先頭

©2009-2025 Movatter.jp