1 Setup

1.1 Packages and options

library(NNbenchmark)
library(kableExtra)
options(scipen = 999)

1.2 Datasets to test

NNdataSummary(NNdatasets)
##      Dataset n_rows n_inputs n_neurons n_parameters
## 1     mDette    500        3         5           26
## 2  mFriedman    500        5         5           36
## 3  mIshigami    500        3        10           51
## 4    mRef153    153        5         3           22
## 5     uDmod1     51        1         6           19
## 6     uDmod2     51        1         5           16
## 7  uDreyfus1     51        1         3           10
## 8  uDreyfus2     51        1         3           10
## 9    uGauss1    250        1         5           16
## 10   uGauss2    250        1         4           13
## 11   uGauss3    250        1         4           13
## 12 uNeuroOne     51        1         2            7

1.3 rminer trainPredict arguments - inputs formula, data

if(dir.exists("D:/GSoC2020/Results/2020run03/"))
{  
  odir <- "D:/GSoC2020/Results/2020run03/"
}else if(dir.exists("~/Documents/recherche-enseignement/code/R/NNbenchmark-project/NNtempresult/"))
{  
  odir <- "~/Documents/recherche-enseignement/code/R/NNbenchmark-project/NNtempresult/"
}else
  odir <- "~"

nrep <- 10
maxit2ndorder  <-    200
maxit1storderA <-   1000
maxit1storderB <-  10000
maxit1storderC <- 100000

#library(rminer)
rminer.method <- "none"
hyperParams.rminer <- function(...) {
    return (list(task="reg", iter=maxit2ndorder))
}
NNtrain.rminer <- function(x, y, dataxy, formula, neur, method, hyperParams, ...) {
    
    hyper_params <- do.call(hyperParams, list(...))
    
    rminer::fit(formula, data = dataxy, model = "mlp", task = hyper_params$task, 
                                        size = neur, maxit = hyper_params$iter)
}
NNpredict.rminer <- function(object, x, ...)
   as.numeric(rminer::predict(object, newdata=as.data.frame(x)))
NNclose.rminer <- function()
  if("package:rminer" %in% search())
    detach("package:rminer", unload=TRUE)
rminer.prepareZZ <- list(xdmv = "m", ydmv = "m", zdm = "d", scale = TRUE)
if(FALSE)
res <- trainPredict_1data(1, rminer.method, "NNtrain.rminer", "hyperParams.rminer", "NNpredict.rminer", 
                               NNsummary, "NNclose.rminer", NA, rminer.prepareZZ, nrep=2,
                               echo=TRUE, doplot=FALSE, echoreport=0,
                               pkgname="rminer", pkgfun="fit", csvfile=TRUE, rdafile=TRUE, odir=odir)

2 Launch package’s trainPredict

res <- trainPredict_1pkg(1:12, pkgname = "rminer", pkgfun = "fit", rminer.method,
  prepareZZ.arg = rminer.prepareZZ, nrep = nrep, doplot = TRUE,
  csvfile = TRUE, rdafile = TRUE, odir = odir, echo = FALSE)

3 Results

#print(res)
kable(t(apply(res, c(1,4), min)))%>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed"))
RMSE MSE MAE WAE time
mDette 0.2344 0.0549 0.1877 0.7583 0.21
mFriedman 0.0095 0.0001 0.0074 0.0290 0.26
mIshigami 0.5619 0.3157 0.4061 2.4458 0.43
mRef153 3.1782 10.1006 2.1846 12.9657 0.03
uDmod1 0.0413 0.0017 0.0322 0.1104 0.01
uDmod2 0.0407 0.0017 0.0324 0.0920 0.01
uDreyfus1 0.0015 0.0000 0.0013 0.0030 0.00
uDreyfus2 0.0906 0.0082 0.0721 0.2140 0.00
uGauss1 2.2370 5.0043 1.7337 6.9997 0.08
uGauss2 2.3543 5.5428 1.8499 7.4454 0.05
uGauss3 2.2827 5.2109 1.8118 6.3302 0.06
uNeuroOne 0.2830 0.0801 0.2313 0.5675 0.00 
kable(t(apply(res, c(1,4), median)))%>%
  kable_styling(bootstrap_options = c("striped", "hover", "condensed"))
RMSE MSE MAE WAE time
mDette 0.30460 0.09285 0.23550 1.34855 0.250
mFriedman 0.01060 0.00010 0.00835 0.03765 0.285
mIshigami 0.65935 0.43470 0.50710 2.86255 0.440
mRef153 3.24020 10.49915 2.26760 13.50975 0.035
uDmod1 0.04400 0.00190 0.03550 0.12090 0.025
uDmod2 0.05005 0.00255 0.04110 0.11065 0.025
uDreyfus1 0.00230 0.00000 0.00195 0.00730 0.010
uDreyfus2 0.09060 0.00820 0.07245 0.22000 0.020
uGauss1 2.31220 5.34635 1.81320 7.43790 0.095
uGauss2 2.37565 5.64385 1.86935 7.60370 0.075
uGauss3 2.77235 7.68820 2.18500 7.50600 0.070
uNeuroOne 0.28300 0.08010 0.23130 0.56750 0.010