What can we learn about the psychiatric diagnostic categories by analysing patients' lived experiences with Machine-Learning?

Table 2 Test of similarity of narratives obtained within and across clinical diagnoses

	Cluster 1	Cluster 2	Cluster 3
Cluster 0	0.33	0.43	0.42
Cluster 0	(Common conditions: auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness)	(Common conditions: auditory hallucination, depressed mood and loss of interest, fear, feeling sick, pain, sadness)	(Common conditions: auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness, sleep)
Cluster 1		0.43	0.54
Cluster 1		(Common conditions: anxiety, auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness)	(Common conditions: anxiety, attention deficit, auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness)
Cluster 2	0.43		0.43
Cluster 2	(Common conditions: anxiety, auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness)		(Common conditions: anxiety, auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness)
Cluster 3	0.43	0.43
Cluster 3	(Common conditions: anxiety, auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness)	(Common conditions: anxiety, auditory hallucination, depressed mood and loss of interest, fear, feeling sick, sadness)

Note. The table depicts the average Jaccard’s coefficient—either with a diagnosis of Cluster 0, 1, 2 and 3

ISSN: 1471-244X