Can XGBoost be applied to time series data to forecast?

forecasting

#1

For discussions related to modeling, machine learning and deep learning. Related packages include caret, modelr, yardstick, rsample, parsnip, tensorflow, keras, cloudml, and tfestimators.

Hi Forecasting experts,

I have temperature variation data from Jan 2014 to Aug 2018. I would like to predict the temperature from Sept 2018 to Sept 2019.

I wanted to try XGBoost but not sure if it captures seasonalities, trend etc
Found an example here (https://www.kaggle.com/furiousx7/xgboost-time-series) but it is in python and also couldnt interpret it correctly as my data has few categorical variable as well.

Also, Can you suggest which forecasting algorithm might work here.
Example would be greatly appreciated.

Thanks in advance


#2

You have to apply xgboost and see whether it can capture the seasonal variations.
The other option is to use ets or Arima models in the forecast package. Both these methods can handle seasonal variations. These methods are applicable to univariate time series.
You can also use neural networks. Good old shallow neural network can produce excellent forecasts.
For including exogenous variables, you may try the arimax function.


closed #3

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.