RMSE: RMSE Calculator Function

In gradDescent: Gradient Descent for Regression Tasks

Description

A function to calculate error using Root-Mean-Square-Error

Usage

RMSE(outputData, prediction)

Arguments

outputData	a data.frame represented dataset with output variable only (m \times 1), where m is the number of instances has one variable, which is the output.
prediction	a data.frame represented prediction data with output variable only (m \times 1), where m is the number of instances has one variable, which is the output.

Details

This function used to calculate the error between two variables. outputData is the first parameter of this function representing the real output value. prediction is the second parameter of this function representing the prediction value.

Value

a float value represent the average error of the prediction

Examples

##################################
## Calculate Error using RMSE
## load R Package data
data(gradDescentRData)
## get z-factor Data
dataSet <- gradDescentRData$CompressilbilityFactor
## do variance scaling to dataset
featureScalingResult <- varianceScaling(dataSet)
## split dataset
splitedDataSet <- splitData(featureScalingResult$scaledDataSet)
## built model using GD
model <- GD(splitedDataSet$dataTrain)
## separate testing data with input only
dataTestInput <- (splitedDataSet$dataTest)[,1:ncol(splitedDataSet$dataTest)-1]
## predict testing data using GD model
predictionData <- prediction(model, dataTestInput)
## calculate error using rmse
errorValue <- RMSE(splitedDataSet$dataTest[,ncol(splitedDataSet$dataTest)],
predictionData[,ncol(predictionData)])
## show result
errorValue

gradDescent documentation built on May 2, 2019, 9:42 a.m.