Time series data prediction

nlp
1

Hallo Everyone!

Apology in advance since I am not sure this is the right place to ask such question.
I have some IMU sensor data. These data describes what individual players did e.g.
Jogging, running, long pass, short pass etc. I want to develop a model using this data.

The target is to predict whether the session was a training one or a real match or something else.
The first idea comes to my mind is to use LSTM or any special version of LSTM which will be able to predict the Event type (training , match etc.) from given activities(e.g. running, pass, jogging) after the end of the session.

Any idea would be really appreciated.

2

Can you describe the data in more detail, or show paste some of it.

3

time label.ball_contact label.movement_speed label.movement_distance event.kick.confidence event.kick.ball_speed event.motion.speed.average event.motion.speed.peak event.motion.distance challenge.idx challenge.id Event
0 1.372256 0.0 -1.000000 -1.0 0.00 0.000000 0.000 0.000 0.000000 0.0 0.0 No_Event
…
1000 1.472256 0.0 -1.000000 -1.0 0.00 0.000000 0.000 0.000 0.000000 0.0 0.0 Match
…

200 1.572256 0.0 -1.000000 -1.0 0.00 0.000000 0.000 0.000 0.000000 0.0 0.0 Exercise
…