#At first the program should read the .txt file with the points that is going to be trainned and the other one that is going to be use to make the prediction. In this example is used the file with the sections spaced 500 meters. #PS:The user should creat a project and put the txt files that are available in the same file as the project
train<-read.table("./train500.txt", sep = ";", header = TRUE) prediction<-read.table("./pred.txt", sep = ";", header = TRUE)
#Trainning the ANN and printing print(nn500 <- neuralnet(Z_Normalized ~ RED_N + NIR_N, data=train, hidden=3, rep=100, err.fct="sse", linear.output=TRUE))
plot(nn500)
#Doing the prediction for the "prediction" points pred_ANN_500 <- predict(nn500,"prediction")
#Exporting the results for a .txt file
write.table(pred_ANN_500, file = "pred_ANN_500.txt", row.names = TRUE, sep = ";", dec = ",")