will it work for multivariate time series

[Sandy4321]

great code thanks may you clarify : will it work for multivariate time series prediction both regression and classification 1 where all values are continues values 2 or even will it work for multivariate time series where values are mixture of continues and categorical values for example 2 dimensions have continues values and 3 dimensions are categorical values

color        weight     gender  height  age  

1 black 56 m 160 34 2 white 77 f 170 54 3 yellow 87 m 167 43 4 white 55 m 198 72 5 white 88 f 176 32

[Charly52830]

Hi Sandy, here are some important points to take care in:

1. The prediction methods are for predicting 5 steps to the future of multiple univariate time series of continuous data (around 4035 time series which represents the number of students enrolled each year in schools), however, here is nothing related to time series classification.
2. You can easily augment the number of steps to more than 5, but I recommend to adjust the hyperparameters of each method.
3. The methods are not intended for multivariate time series. You could apply these methods by taking individually the series from the multivariate time series, however, this would not take the advantage of the multivariate component in which, for example, you could take predictors from two different but correlated time series for predicting one or another time series.
4. The methods are not intended for categorical data. I think an interesting approach for this would be to apply Fuzzy time series methods or Fuzzy logic to transform linguistic variables to numerical variables if possible. For more about Fuzzy time series here is a good tutorial from Petrônio Silva: https://towardsdatascience.com/a-short-tutorial-on-fuzzy-time-series-dcc6d4eb1b15

I hope your questions were answered :)