ttsLSTM: Train time series by LSTM of 'tensorflow' provided by 'kera'

In iForecast: Machine Learning Time Series Forecasting

Train time series by LSTM of tensorflow provided by kera

Description

It generates both the static and recursive time series plots of deep learning LSTM object generated by package tensorflow provided by kera.

Usage

ttsLSTM(y,
  x=NULL,
  train.end,
  arOrder=1,
  xregOrder=0,
  type,
  memoryLoops=10,
  shape=NULL,
  dim3=5,
  batch.range=2:7,
  batch.size=NULL)

Arguments

y	The time series object of the target variable, or the dependent variable, with timeSeries or zoo format, must have dimension. y can be both continuous and discrete.Time format must be "
x	The time series matrix of input variables, or the independent variables, with timeSeries or zoo format. Time format must be "
train.end	The end date of training data, must be specificed.The default dates of train.start and test.end are the start and the end of input data; and the test.start is the 1-period next of train.end.
arOrder	The autoregressive order of the target variable, which may be sequentially specifed like arOrder=1:5; or discontinuous lags like arOrder=c(1,3,5); zero is not allowed.Default is 1.
xregOrder	The distributed lag structure of the input variables, which may be sequentially specifed like xregOrder=1:5; or discontinuous lags like xregOrder=c(0,3,5); zero is allowed since contemporaneous correlation is allowed.
type	The additional input variables. We have four selection: "none"=no other variables, "trend"=inclusion of time dummy, "season"=inclusion of seasonal dummies, "both"=inclusion of both trend and season. No default.
memoryLoops	Length of LSTM learning network loop, to achieve better learning results, this not is suggested to be the same as the length of data row. Default is 10.

.

shape

The second dmension of LSTM array. If NULL, then it will use the number of columns of complete dataset.

.

dim3	The third dmension of LSTM array. Default is 5.

.

batch.range

The range of search batch.size. The code selects the first that satisfies exact division with the rows of data used

.

batch.size

The number of batch size for LSTM layer. Default is NULL determined by searching among the batch.range.

.

Details

This function calls the function fit of package tensorflow to execute Long-Short Term Memory (LSTM) estimation. When execution finished, it computes two types of time series forecasts: static and recursive.

Value

output	Output object generated by train function of caret.
batch.size	The batch.size used for LSTM network.
k	The third dimension of arrayin LSTM network.
SHAPE	The shape size of array in LSTM network.
arOrder	he autoregressive order of the target variable used.
data	The dataset of used.
dataused	The data used by arOrder, xregOrder, and type

Author(s)

Ho Tsung-wu <[e-mail]>, College of Management, National Taiwan Normal University.

Examples

# Cross-validation takes time, example below is commented.
data("macrodata")
dep<-macrodata[,"unrate",drop=FALSE]
ind<-macrodata[,-1,drop=FALSE]

# Choosing the dates of training and testing data
train.end<-"2008-12-01"


#RNN with LSTM network
#LSTM<-ttsLSTM(y=dep, x=ind, train.end,arOrder=c(2,4), xregOrder=c(1,4),
# memoryLoops=5, type=c("none","trend","season","both")[4],
# batch.range=2:7,batch.size=NULL)

#testData3<-window(LSTM$dataused,start="2009-01-01",end=end(LSTM$data))
#P1<-iForecast(Model=LSTM,newdata=testData3,type="static")
#P2<-iForecast(Model=LSTM,newdata=testData3,type="dynamic")

#tail(cbind(testData3[,1],P1,P2))

iForecast documentation built on July 10, 2023, 1:59 a.m.