Table 1.

Table summarizing the datasets from the six subjects used to train the neural network and details of the validation dataset. Each row represents one separate dataset, and the columns indicate how many cycles belong to each class.

Dataset	DIA	DK	DP	TCK	HBN	TRN	tDIA	fDIA
Training Data:
Sub. 1 Granåsen/lab	266	79	433	0	0	0	0	0
Sub. 1 Holmenkollen 1	255	85	201	0	0	0	0	4
Sub. 1 Holmenkollen 2	0	0	22	0	0	0	0	0
Sub. 1 Holmenkollen 3	0	50	0	0	0	0	0	0
Sub. 1 Holmenkollen 4	70	0	0	0	0	0	0	0
Sub. 1 Natrudstilen 1	80	22	54	4	10	13	3	3
Sub. 1 Natrudstilen 2	82	15	59	6	7	8	3	4
Sub. 1 Natrudstilen 3	48	1	54	1	11	11	0	2
Sub. 3 Natrudstilen 1	29	7	39	2	10	15	0	0
Sub. 4 Natrudstilen	33	15	46	4	0	14	2	0
Sub. 5 Sjusjøen	256	176	124	12	0	0	17	14
Sub. 9 Natrudstilen	394	432	428	0	0	0	18	11
Sub. 10 Natrudstilen	96	61	84	0	0	0	0	0
Percentage of dataset	37.4%	21.9%	35.8%	0.7%	0.9%	1.4%	1%	0.9%
Validation dataset:
Sub 3. Natrudstilen 2	20	2	86	3	5	22	2	0
Percentage of dataset	14%	2%	61%	2%	4%	16%	1%	0%