Can I use RBF neural network to forecast time series in R language? -


there time series of number of jobs in manufacture between 1978 2017. want use radial basis neural network forecast number of jobs in 2 years. possible? if is, please write code in r language? many thanks! wrote codes here :

install.packages("rsnns")  library(rsnns)  data <- read.csv("jobs.csv",header = true)  tsa01 <- ts(data$`a-01`,start = c(1978,2),end = c(2017,1),frequency = 4) 

part of data shown in image below:

enter image description here

looking @ example of data, have simple dataset: response variable (number of jobs), , single covariate (date). if limit of data, there no need neural network approach. neural networks , other supervised machine learning approaches necessary when have tons of features (i.e., covariates, called "p"), typically such p >> n (number of observations). in specific case, start simple linear regression perhaps takes things month or season account covariates. if regression looks good, can make predictions future time points.

if have more complex data eluded in question, there great textbook machine learning available online free. includes number of laboratory chapters written in r guide through various analyses, invest time in reading pros , cons of various approaches before decide use neural networks specifically. can find textbook here: http://www-bcf.usc.edu/~gareth/isl/ (and click "download book pdf".


Comments

Popular posts from this blog

node.js - Node js - Trying to send POST request, but it is not loading javascript content -

javascript - Replicate keyboard event with html button -

javascript - Web audio api 5.1 surround example not working in firefox -