## Actimetric data segmentation and classification for patients daily activities prediction.

In the context of post-stroke patients rehabilitation, when they return home the CHU of Montpellier provides patients with accelerometric bracelets which they send back to the CHU after one week, together with a completed activity diary. This helps CHU therapists (physiotherapists, medical doctors, etc.) to monitor the patients activities at home. They are especially interested in some physical activities that involves both arms. The analysis of these key activities can indeed highlight the compensation of the paretic (i.e. damaged) arm by the non-paretic (i.e. healthy) one.

In this competition you have to predict the activities performed by a given patient during a given (6th) day between 7am and 19pm of a 1 week actimetric recording from:
- the patient actimetric data of the 6 first days
- the patient activities journal of the 5 first days