We provide a forecasting model for time series forecasting problems with predictors. The offered model, which is based on a submitted research and called tree-based moving average (TBMA), is based on the integration of the moving average approach to tree-based ensemble approach. The tree-based ensemble models can capture the complex correlations between the predictors and response variable but lack in modelling time series components. The integration of the moving average approach to the tree-based ensemble approach helps the TBMA model to handle both correlations and autocorrelations in time series data. This package provides a tbma() forecasting function that utilizes the ranger() function from the 'ranger' package. With the help of the ranger() function, various types of tree-based ensemble models, such as extremely randomized trees and random forests, can be used in the TBMA model.
A new tree-based ensemble model that is called tree-based moving average (TBMA) is provided for time series forecasting problems. The TBMA model provides point and probabilistic forecasts and uses both of the tree-based ensemble and MA approaches to consider predictors and time series components. The suggested model can handle a large number of numerical and categorical predictors without sacrificing the accuracy with the help of the tree-based ensemble approach. The integration of the MA model to a tree-based ensemble approach helps the TBMA model to capture autocorrelation between observations. With the use of the tree-based ensemble and MA approaches, the proposed model becomes multivariate and adapted to time series problems.
You can install the released version of tbma from GitHub with:
library(devtools)
install_github("BurakhanS/tbma")
library(tbma)
This is a basic example which shows you how to solve a common problem:
library(datasets)
library(data.table)
library(tbma)
data(airquality)
summary(airquality)
#> Ozone Solar.R Wind Temp
#> Min. : 1.00 Min. : 7.0 Min. : 1.700 Min. :56.00
#> 1st Qu.: 18.00 1st Qu.:115.8 1st Qu.: 7.400 1st Qu.:72.00
#> Median : 31.50 Median :205.0 Median : 9.700 Median :79.00
#> Mean : 42.13 Mean :185.9 Mean : 9.958 Mean :77.88
#> 3rd Qu.: 63.25 3rd Qu.:258.8 3rd Qu.:11.500 3rd Qu.:85.00
#> Max. :168.00 Max. :334.0 Max. :20.700 Max. :97.00
#> NA's :37 NA's :7
#> Month Day
#> Min. :5.000 Min. : 1.0
#> 1st Qu.:6.000 1st Qu.: 8.0
#> Median :7.000 Median :16.0
#> Mean :6.993 Mean :15.8
#> 3rd Qu.:8.000 3rd Qu.:23.0
#> Max. :9.000 Max. :31.0
#>
airquality<-as.data.table(airquality)
airquality[complete.cases(airquality)]
#> Ozone Solar.R Wind Temp Month Day
#> 1: 41 190 7.4 67 5 1
#> 2: 36 118 8.0 72 5 2
#> 3: 12 149 12.6 74 5 3
#> 4: 18 313 11.5 62 5 4
#> 5: 23 299 8.6 65 5 7
#> ---
#> 107: 14 20 16.6 63 9 25
#> 108: 30 193 6.9 70 9 26
#> 109: 14 191 14.3 75 9 28
#> 110: 18 131 8.0 76 9 29
#> 111: 20 223 11.5 68 9 30
train <- airquality[1:102,]
test <- airquality[103:nrow(airquality), ]
output<-tbma(Temp ~ .,train = train,test = test,
prediction_type = "point",horizon=100,ma_order = 2)
library(datasets)
#this is the input data
head(airquality,5)
#> Ozone Solar.R Wind Temp Month Day
#> 1: 41 190 7.4 67 5 1
#> 2: 36 118 8.0 72 5 2
#> 3: 12 149 12.6 74 5 3
#> 4: 18 313 11.5 62 5 4
#> 5: NA NA 14.3 56 5 5
#this is the output data (Order and prediction columns are added by the tbma() function)
head(output,5)
#> Order Temp Ozone Solar.R Wind Month Day prediction
#> 1: 66 86 44 192 11.5 8 12 83.5
#> 2: 67 82 28 273 11.5 8 13 81.5
#> 3: 68 80 65 157 9.7 8 14 85.0
#> 4: 69 77 22 71 10.3 8 16 81.5
#> 5: 70 79 59 51 6.3 8 17 83.5
``