Learn R Programming

⚠️There's a newer version (3.4.8) of this package.Take me there.

FRESA.CAD (version 3.3.0)

Feature Selection Algorithms for Computer Aided Diagnosis

Description

Contains a set of utilities for building and testing statistical models (linear, logistic,ordinal or COX) for Computer Aided Diagnosis/Prognosis applications. Utilities include data adjustment, univariate analysis, model building, model-validation, longitudinal analysis, reporting and visualization.

Copy Link

Version

Install

install.packages('FRESA.CAD')

Monthly Downloads

309

Version

3.3.0

License

LGPL (>= 2)

Maintainer

Jose Tamez-Pena

Last Published

September 29th, 2020

Functions in FRESA.CAD (3.3.0)

Cross-validated Signature

Data frame normalization

Hybrid Hierarchical Modeling

Automated model selection

EmpiricalSurvDiff

Estimate the LR value and its associated p-values

FilterUnivariate

Univariate Filters

BSWiMS model selection

Data frame used in several examples of this package

FRESA.CAD-package

FeatuRE Selection Algorithms for Computer-Aided Diagnosis (FRESA.CAD)

Ridge Linear Models

ForwardSelection.Model.Res

NeRI-based feature selection procedure for linear, logistic, or Cox proportional hazards regression models

Set Clustering using the Generalized Minimum Volume Ellipsoid (GMVE)

Hybrid Hierarchical Modeling with GMVE and BSWiMS

Analysis of the effect of each term of a binary classification model by analysing its reclassification performance

mRMR.classic_FRESA

FRESA.CAD wrapper of mRMRe::mRMR.classic

Analysis of the effect of each term of a linear regression model by analysing its residuals

Jaccard Index of two labeled sets

listTopCorrelatedVariables

List the variables that are highly correlated with each other

bootstrapValidation_Bin

Bootstrap validation of binary classification models

Cluster Clustering using the Isodata Approach

crossValidationFeatureSelection_Bin

IDI/NRI-based selection of a linear, logistic, or Cox proportional hazards regression model from a set of candidate variables

ForwardSelection.Model.Bin

IDI/NRI-based feature selection procedure for linear, logistic, and Cox proportional hazards regression models

bootstrapValidation_Res

Bootstrap validation of regression models

nearestCentroid

Class Label Based on the Minimum Mahalanobis Distance

Fit a model to the data

predict.FRESA_NAIVEBAYES

Predicts NAIVE_BAYES models

predict.FRESAKNN

Predicts class::knn models

Latent class based modeling of binary outcomes

Naive Bayes Modeling

backVarElimination_Res

NeRI-based backwards variable elimination

bootstrapVarElimination_Bin

IDI/NRI-based backwards variable elimination with bootstrapping

KNN Setup for KNN prediction

reportEquivalentVariables

Report the set of variables that will perform an equivalent IDI discriminant function

rankInverseNormalDataFrame

rank-based inverse normal transformation of the data

predict.FRESA_HLCM

Predicts BOOST_BSWiMS models

predict.FRESAsignature

Predicts CVsignature models

backVarElimination_Bin

IDI/NRI-based backwards variable elimination

plot.bootstrapValidation_Bin

Plot ROC curves of bootstrap results

plot.bootstrapValidation_Res

Plot ROC curves of bootstrap results

predict.FRESA_SVM

Predicts TUNED_SVM models

bootstrapVarElimination_Res

NeRI-based backwards variable elimination with bootstrapping

featureAdjustment

Adjust each listed variable to the provided set of covariates

updateModel.Res

Update the NeRI-based model using new data or new threshold values

updateModel.Bin

Update the IDI/NRI-based model using new data or new threshold values

Get the bagged model from a list of models

A generic fit method with a filtered step for feature selection

improvedResiduals

Estimate the significance of the reduction of predicted residuals

Plot a heat map of selected variables

predict.FRESA_RIDGE

Predicts LM_RIDGE_MIN models

residualForFRESA

Return residuals from prediction

ensemblePredict

The median prediction from a list of models

crossValidationFeatureSelection_Res

NeRI-based selection of a linear, logistic, or Cox proportional hazards regression model from a set of candidate variables

nearestNeighborImpute

nearest neighbor NA imputation

GLMNET fit with feature selection"

Bar plot with error bars

Compare performance of different model fitting/filtering algorithms

signatureDistance

Distance to the signature template

plot.FRESA_benchmark

Plot the results of the model selection benchmark

trajectoriesPolyFeatures

Extract the per patient polynomial Coefficients of a feature trayectory

predict.FRESA_BESS

Predicts BESS models

predict.fitFRESA

Linear or probabilistic prediction

Univariate analysis of features (additional values returned)

getKNNpredictionFromFormula

Predict classification using KNN

predict.FRESA_FILTERFIT

Predicts filteredFit models

Plot test ROC curves of each cross-validation model

predict.CLUSTER_CLASS

Predicts ClustClass outcome

Returns a CV signature template

Report the univariate analysis, the cross-validation analysis and the correlation analysis

predictionStats

Prediction Evaluation

timeSerieAnalysis

Fit the listed time series variables to a given model

Predicts GMVECluster clusters

predict.GMVE_BSWiMS

Predicts GMVEBSWiMS outcome

predict.FRESA_GLMNET

Predicts GLMNET fitted objects

summary.fitFRESA

Returns the summary of the fit

summary.bootstrapValidation_Bin

Generate a report of the results obtained using the bootstrapValidation_Bin function

Cross Validation of Prediction Models

update.uniRankVar

Update the univariate analysis using new data

univariateRankVariables

Univariate analysis of features