Título da msg explica

> sessionInfo()

R version 3.2.3 (2015-12-10)

Platform: x86_64-w64-mingw32/x64 (64-bit)

Running under: Windows >= 8 x64 (build 9200)

locale:

[1] LC_COLLATE=Portuguese_Brazil.1252  LC_CTYPE=Portuguese_Brazil.1252  

[3] LC_MONETARY=Portuguese_Brazil.1252 LC_NUMERIC=C                     

[5] LC_TIME=Portuguese_Brazil.1252   

attached base packages:

[1] stats     graphics  grDevices utils     datasets  methods   base    

other attached packages:

[1] nnet_7.3-11         caret_6.0-64        ggplot2_2.0.0.9001  lattice_0.20-33   

[5] mlbench_2.1-1       RevoUtilsMath_3.2.3

loaded via a namespace (and not attached):

[1] Rcpp_0.12.3         compiler_3.2.3      nloptr_1.0.4        plyr_1.8.3        

 [5] iterators_1.0.8     tools_3.2.3         digest_0.6.9        lme4_1.1-10       

 [9] nlme_3.1-124        gtable_0.1.2        mgcv_1.8-10         Matrix_1.2-3      

[13] foreach_1.4.3       yaml_2.1.13         parallel_3.2.3      SparseM_1.7       

[17] stringr_1.0.0.9000  knitr_1.12          MatrixModels_0.4-2  stats4_3.2.3      

[21] grid_3.2.3          rmarkdown_0.9.4     minqa_1.2.4         reshape2_1.4.1.9000

[25] car_2.1-1           magrittr_1.5        scales_0.3.0        codetools_0.2-14  

[29] htmltools_0.3       MASS_7.3-45         splines_3.2.3       rsconnect_0.4.1.4 

[33] pbkrtest_0.4-5      colorspace_1.2-6    quantreg_5.19       stringi_1.0-1     

[37] munsell_0.4.2     

>

____ código

---

title: "Neural Network example"

author: "Leonard Mendonça de Assis"

date: "21 de janeiro de 2016"

output:

  pdf_document:

    highlight: pygments

    number_sections: yes

  html_document:

    highlight: pygments

    number_sections: yes

  word_document:

    highlight: pygments

---

# Session start

```{r}

require(MASS)

require(neuralnet)

require(plyr)

require(boot)

require(knitr)

data <- Boston

kable(head(data))

kable(

  apply(data,2,function(x) sum(is.na(x))),

  caption='NAs per variables'

)

```

# Adjusting a linear model

```{r}

set.seed(500)

index <- sample(1:nrow(data),round(0.75*nrow(data)))

train <- data[index,]

test <- data[-index,]

lm.fit <- glm(medv~., data=train)

summary(lm.fit)

pr.lm <- predict(lm.fit,test)

MSE.lm <- sum((pr.lm - test$medv)^2)/nrow(test)

```

# Adjusting a Neural Net

As a first step, we are going to address data preprocessing. It is good practice to normalize your data before training a neural network. I cannot emphasize enough how important this step is: depending on your dataset, avoiding normalization may lead to useless results or to a very difficult training process (most of the times the algorithm will not converge before the number of maximum iterations allowed). You can choose different methods to scale the data (z-normalization, min-max scale, etc…). I chose to use the min-max method and scale the data in the interval [0,1].

Usually scaling in the intervals [0,1] or [-1,1] tends to give better results. We therefore scale and split the data before moving on:

```{r}

maxs <- apply(data, 2, max)

mins <- apply(data, 2, min)

scaled <- as.data.frame(scale(data, center = mins, scale = maxs - mins))

train_ <- scaled[index,]

test_ <- scaled[-index,]

```

Note that scale returns a matrix that needs to be coerced into a data.frame.

```{r}

n <- names(train_)

f <- as.formula(paste("medv ~", paste(n[!n %in% "medv"], collapse = " + ")))

nn <- neuralnet(f,data=train_,hidden=c(5,3),linear.output=T)

```

And the Neural Nework looks like this:

```{r}

plot(nn)

```

```{r}

sessionInfo()

```