Scribd
Upload
0 views

14.3 Neural Networks

The document outlines the process of building a neural network model to predict the strength of concrete using various input features. It includes data normalization, model training, and performance evaluation, showing improvements in correlation with the addition of hidden nodes. The final model achieved a correlation of approximately 0.94 with the test data, indicating enhanced predictive accuracy.

Uploaded by

resiro9407
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
0 views

14.3 Neural Networks

The document outlines the process of building a neural network model to predict the strength of concrete using various input features. It includes data normalization, model training, and performance evaluation, showing improvements in correlation with the addition of hidden nodes. The final model achieved a correlation of approximately 0.94 with the test data, indicating enhanced predictive accuracy.

Uploaded by

resiro9407
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 6

Prepare a model for strength of concrete data using Neural Networks

Ans:

> library(neuralnet)
> library(NeuralNetTools)
> library(nnet)
> library(caret)
> library(corrplot)
> library(GGally)

> concrete <- read.csv(file.choose())


> head(concrete)
cement slag ash water superplastic coarseagg fineagg age strength
1 141.3 212.0 0.0 203.5 0.0 971.8 748.5 28 29.89
2 168.9 42.2 124.3 158.3 10.8 1080.8 796.2 14 23.51
3 250.0 0.0 95.7 187.4 5.5 956.9 861.2 28 29.22
4 266.0 114.0 0.0 228.0 0.0 932.0 670.0 28 45.85
5 154.8 183.4 0.0 193.3 9.1 1047.4 696.7 28 18.29
6 255.0 0.0 0.0 192.0 0.0 889.8 945.0 90 21.86

> str(concrete)
'data.frame': 1030 obs. of 9 variables:
$ cement : num 141 169 250 266 155 ...
$ slag : num 212 42.2 0 114 183.4 ...
$ ash : num 0 124.3 95.7 0 0 ...
$ water : num 204 158 187 228 193 ...
$ superplastic: num 0 10.8 5.5 0 9.1 0 0 6.4 0 9 ...
$ coarseagg : num 972 1081 957 932 1047 ...
$ fineagg : num 748 796 861 670 697 ...
$ age : int 28 14 28 28 28 90 7 56 28 28 ...
$ strength : num 29.9 23.5 29.2 45.9 18.3 ...

> summary(concrete)
cement slag ash water superplastic coarseagg
Min. :102.0 Min. : 0.0 Min. : 0.00 Min. :121.8 Min. : 0.000 Min. : 801.0
1st Qu.:192.4 1st Qu.: 0.0 1st Qu.: 0.00 1st Qu.:164.9 1st Qu.: 0.000 1st Qu.: 932.0
Median :272.9 Median : 22.0 Median : 0.00 Median :185.0 Median : 6.400 Median : 968.0
Mean :281.2 Mean : 73.9 Mean : 54.19 Mean :181.6 Mean : 6.205 Mean : 972.9
3rd Qu.:350.0 3rd Qu.:142.9 3rd Qu.:118.30 3rd Qu.:192.0 3rd Qu.:10.200 3rd Qu.:1029.4
Max. :540.0 Max. :359.4 Max. :200.10 Max. :247.0 Max. :32.200 Max. :1145.0
fineagg age strength
Min. :594.0 Min. : 1.00 Min. : 2.33
1st Qu.:731.0 1st Qu.: 7.00 1st Qu.:23.71
Median :779.5 Median : 28.00 Median :34.45
Mean :773.6 Mean : 45.66 Mean :35.82
3rd Qu.:824.0 3rd Qu.: 56.00 3rd Qu.:46.13
Max. :992.6 Max. :365.00 Max. :82.60
> attach(concrete)

Normalising function

> normm <- function(x){


return((x-min(x))/(max(x)-min(x)))
}

Normalising df
> norm_concrete <- as.data.frame(lapply(concrete,normm))

> head(norm_concrete)
cement slag ash water superplastic coarseagg fineagg
1 0.08972603 0.5898720 0.0000000 0.6525559 0.0000000 0.4965116 0.3876066
2 0.15273973 0.1174179 0.6211894 0.2915335 0.3354037 0.8133721 0.5072755
3 0.33789954 0.0000000 0.4782609 0.5239617 0.1708075 0.4531977 0.6703462
4 0.37442922 0.3171953 0.0000000 0.8482428 0.0000000 0.3808140 0.1906673
5 0.12054795 0.5102949 0.0000000 0.5710863 0.2826087 0.7162791 0.2576518
6 0.34931507 0.0000000 0.0000000 0.5607029 0.0000000 0.2581395 0.8805820
age strength
1 0.07417582 0.3433412
2 0.03571429 0.2638595
3 0.07417582 0.3349944
4 0.07417582 0.5421702
5 0.07417582 0.1988290
6 0.24450549 0.2433038

Splitting of data to test and train

> trainn <- createDataPartition(norm_concrete$strength,p=0.8,list = F)


> trn_c <- norm_concrete[trainn,]
> tst_c <- norm_concrete[-trainn,]

Model building

> model_con <- neuralnet(strength~.,data = trn_c)


> str(model_con)
List of 14
$ call : language neuralnet(formula = strength ~ ., data = trn_c)
$ response : num [1:826, 1] 0.343 0.264 0.335 0.542 0.199 ...
..- attr(*, "dimnames")=List of 2
.. ..$ : chr [1:826] "1" "2" "3" "4" ...
.. ..$ : chr "strength"
$ covariate : num [1:826, 1:8] 0.0897 0.1527 0.3379 0.3744 0.1205 ...
..- attr(*, "dimnames")=List of 2
.. ..$ : chr [1:826] "1" "2" "3" "4" ...
.. ..$ : chr [1:8] "cement" "slag" "ash" "water" ...
$ model.list :List of 2
..$ response : chr "strength"
..$ variables: chr [1:8] "cement" "slag" "ash" "water" ...
$ err.fct :function (x, y)
..- attr(*, "type")= chr "sse"
$ act.fct :function (x)
..- attr(*, "type")= chr "logistic"
$ linear.output : logi TRUE
$ data :'data.frame': 826 obs. of 9 variables:
..$ cement : num [1:826] 0.0897 0.1527 0.3379 0.3744 0.1205 ...
..$ slag : num [1:826] 0.59 0.117 0 0.317 0.51 ...
..$ ash : num [1:826] 0 0.621 0.478 0 0 ...
..$ water : num [1:826] 0.653 0.292 0.524 0.848 0.571 ...
..$ superplastic: num [1:826] 0 0.335 0.171 0 0.283 ...
..$ coarseagg : num [1:826] 0.497 0.813 0.453 0.381 0.716 ...
..$ fineagg : num [1:826] 0.388 0.507 0.67 0.191 0.258 ...
..$ age : num [1:826] 0.0742 0.0357 0.0742 0.0742 0.0742 ...
..$ strength : num [1:826] 0.343 0.264 0.335 0.542 0.199 ...
$ exclude : NULL
$ net.result :List of 1
..$ : num [1:826, 1] 0.219 0.247 0.281 0.253 0.291 ...
.. ..- attr(*, "dimnames")=List of 2
.. .. ..$ : chr [1:826] "1" "2" "3" "4" ...
.. .. ..$ : NULL
$ weights :List of 1
..$ :List of 2
.. ..$ : num [1:9, 1] 1.747 -3.991 -2.181 -0.617 1.987 ...
.. ..$ : num [1:2, 1] 0.67 -0.678
$ result.matrix : num [1:14, 1] 5.55 9.01e-03 1.72e+03 1.75 -3.99 ...
..- attr(*, "dimnames")=List of 2
.. ..$ : chr [1:14] "error" "reached.threshold" "steps" "Intercept.to.1layhid1" ...
.. ..$ : NULL
- attr(*, "class")= chr "nn"

> plot(model_con)
> pred_model_con <-compute(model_con,tst_c)
> predict_con <- pred_model_con$net.result
> cor(predict_con,tst_c$strength)
[,1]
[1,] 0.8219754
SSE is more with less acc

> plotnet(model_con,cex=0.8)

Improving model performance by including hidden nodes

> model2_con <- neuralnet(strength~.,data = trn_c,hidden = 5)


> str(model2_con)
List of 14
$ call : language neuralnet(formula = strength ~ ., data = trn_c, hidden = 5)
$ response : num [1:826, 1] 0.343 0.264 0.335 0.542 0.199 ...
..- attr(*, "dimnames")=List of 2
.. ..$ : chr [1:826] "1" "2" "3" "4" ...
.. ..$ : chr "strength"
$ covariate : num [1:826, 1:8] 0.0897 0.1527 0.3379 0.3744 0.1205 ...
..- attr(*, "dimnames")=List of 2
.. ..$ : chr [1:826] "1" "2" "3" "4" ...
.. ..$ : chr [1:8] "cement" "slag" "ash" "water" ...
$ model.list :List of 2
..$ response : chr "strength"
..$ variables: chr [1:8] "cement" "slag" "ash" "water" ...
$ err.fct :function (x, y)
..- attr(*, "type")= chr "sse"
$ act.fct :function (x)
..- attr(*, "type")= chr "logistic"
$ linear.output : logi TRUE
$ data :'data.frame': 826 obs. of 9 variables:
..$ cement : num [1:826] 0.0897 0.1527 0.3379 0.3744 0.1205 ...
..$ slag : num [1:826] 0.59 0.117 0 0.317 0.51 ...
..$ ash : num [1:826] 0 0.621 0.478 0 0 ...
..$ water : num [1:826] 0.653 0.292 0.524 0.848 0.571 ...
..$ superplastic: num [1:826] 0 0.335 0.171 0 0.283 ...
..$ coarseagg : num [1:826] 0.497 0.813 0.453 0.381 0.716 ...
..$ fineagg : num [1:826] 0.388 0.507 0.67 0.191 0.258 ...
..$ age : num [1:826] 0.0742 0.0357 0.0742 0.0742 0.0742 ...
..$ strength : num [1:826] 0.343 0.264 0.335 0.542 0.199 ...
$ exclude : NULL
$ net.result :List of 1
..$ : num [1:826, 1] 0.276 0.29 0.389 0.526 0.31 ...
.. ..- attr(*, "dimnames")=List of 2
.. .. ..$ : chr [1:826] "1" "2" "3" "4" ...
.. .. ..$ : NULL
$ weights :List of 1
..$ :List of 2
.. ..$ : num [1:9, 1:5] -5.856 2.374 2.628 1.409 0.532 ...
.. ..$ : num [1:6, 1] 1.3826 -3.6579 1.3653 0.0791 -2.3792 ...
$ result.matrix : num [1:54, 1] 1.78 9.46e-03 2.54e+04 -5.86 2.37 ...
..- attr(*, "dimnames")=List of 2
.. ..$ : chr [1:54] "error" "reached.threshold" "steps" "Intercept.to.1layhid1" ...
.. ..$ : NULL
- attr(*, "class")= chr "nn"

> plot(model2_con)
> predict_model2 <- compute(model2_con,tst_c)
> predict_con2 <- predict_model2$net.result
> cor(predict_con2,tst_c$strength)
[,1]
[1,] 0.9383941
SSE has decreased

> plotnet(model2_con,cex=0.8)

You might also like