Identifying and Monitoring the Daily Routine of Seniors Living at Home

. 2022 Jan 27;22(3):992. doi: 10.3390/s22030992

Algorithm 1: Baseline detection considering both activity sequence and activity length

Inputs:

M M

—transition probability matrix,

P_{S}

—set holding each activity probability to be the first one of the day,

P_{E}

—set holding each activity probability to be the last one of the day,

E P

—end probability weight,

L_{M}

—length median,

L D_{M a x}

—length probability weight,

P

—the transition probability from an activity

A_{i}

to an activity

A_{j}

;
Outputs:

B

—activity sequence representing the baseline.
Begin
1

B \leftarrow [];

l a s t V a l \leftarrow m a x (P_{S});

l a s t A c t i v i t y \leftarrow l a b e l (l a s t V a l);

p r e v V a l \leftarrow n u l l, p r e v A c t i v i t y \leftarrow n u l l;

L D \leftarrow L D_{M a x}

6 while

\exists A_{i} s u c h M M [l a s t A c t i v i t y, A_{i}] > P_{E} [l a s t A c t i v i t y] \times E P - L D

do
7

a p p e n d (B, l a s t A c t i v i t y);

p r e v V a l \leftarrow l a s t V a l;

p r e v A c t i v i t y \leftarrow l a s t A c t i v i t y;

L D \leftarrow i n t e r p o l a t e (l e n (B), [0, L_{M}], [L D_{M a x}, 0]);

l a s t V a l \leftarrow P_{E} [p r e v A c t i v i t y] \times E P - L D;

12 foreach

A_{i}

do
13

P \leftarrow M M [p r e v A c t i v i t y, A_{i}]

14 if

l a s t V a l < p r e v V a l \times P

then
15

l a s t V a l \leftarrow p r e v V a l \times T P;

l a s t A c t i v i t y \leftarrow A_{i}

17     end
18     end
19     end
20     return

B

End