will it work for multivariate time series prediction both regression and classification

[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
weight height age target
1 56 160 34 1.2
2 77 170 54 3.5
3 87 167 43 0.7
4 55 198 72 0.5
5 88 176 32 2.3

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  target 

1 black 56 m 160 34 yes
2 white 77 f 170 54 no
3 yellow 87 m 167 43 yes
4 white 55 m 198 72 no
5 white 88 f 176 32 yes

[nils-braun]

Thanks for your question!
I think I did not 100% understand it though.

I assume you refer to the features selection part.

Do you refer to the features or the target when you say continuous or categorical?

tsfresh will work for all kind of combinations (feature real or binary, target real or binary). For multi-class selection there exist options for the features selection (which turn this on).

[Sandy4321]

yes Feature extraction for further use for multivariate time series prediction both regression and classification

do you have example where Feature extraction and after it used for multivariate time series prediction both regression and classification

[nils-braun]

So just to clarify (because I am still confused concerning your example), tsfresh is a time series feature extraction library. It is used to extract features on time series (your data are just "normal" tabular data examples, in which case you probably want to use something similar to https://www.featuretools.com/).
In case you really have a time series use case: yes, we do have examples for prediction with a multivariate method afterwards. For example this, which is a forecasting use case for a regression target. We also have one for classification. And even a more complex one on multiclass feature selection.

[Sandy4321]

in tabular data order of rows is not important
but in my example order of rows is important
consider it like this
each row means the time series, and columns represent different continues or/and discrete/category observation.

1
where all values are continues values
weight height age target
time 1| 56 160 34 1.2
time 2| 77 170 54 3.5
time 3| 87 167 43 0.7
time 4| 55 198 72 0.5
time 5| 88 176 32 2.3
etc

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      target 

time 1| black 56 m 160 34 yes
time 2| white 77 f 170 54 no
time 3| yellow 87 m 167 43 yes
time 4| white 55 m 198 72 no
time 5| white 88 f 176 32 yes

etc

[nils-braun]

Ah, now I understand.

• Yes, tsfresh will work for time series prediction with continous values - both for regression and prediction. Just a note: tsfresh is a feature extraction and selection library. So you would need to train a ML method afterwards using those features (and which method you use also depends, if you can have a regression or classification target)
• No, categorical values are not supported by most of the features tsfresh extracts.

[lynBIT]

Hi, I have some time series features for regression analysis to model the number of interface calls versus the container CPU, the independent variable is the number of interface calls (two features respectively) and the dependent variable is CPU usage. Both independent and dependent variables are time series data of a sample with an interval of 15s. Now I want to use TsFresh to extract features for regression analysis (other machine learning methods can be used for regression), but I don't know if TsFresh can do that.
this is my dataframe
time features_1 feature_2 cpu
2022-02-21 09:30:00 10 12 0.106
2022-02-21 09:30:15 12 14 0.111
2022-02-21 09:30:30 15 16 0.142
2022-02-21 09:30:45 19 20 0.142
2022-02-21 09:31:00 30 31 0.18
2022-02-21 09:31:15 37 23 0.192
2022-02-21 09:31:30 25 36 0.193
2022-02-21 09:31:45 35 37 0.193
2022-02-21 09:32:00 23 39 0.201
2022-02-21 09:32:15 32 45 0.211
2022-02-21 09:32:30 33 47 0.214
2022-02-21 09:32:45 56 48 0.214
2022-02-21 09:33:00 55 49 0.215
2022-02-21 09:33:15 45 36 0.189
2022-02-21 09:33:30 28 36 0.154
2022-02-21 09:33:45 34 33 0.154
2022-02-21 09:34:00 23 27 0.144
2022-02-21 09:34:15 22 24 0.142
2022-02-21 09:34:30 21 18 0.137
2022-02-21 09:34:45 16 15 0.134