Fig. 1: Overall workflow. | Schizophrenia

Fig. 1: Overall workflow.

From: Leveraging computational linguistics and machine learning for detection of ultra-high risk of mental health disorders in youths

Fig. 1

All HiSoC (High-risk Social Challenge) video recordings were manually transcribed. Feature extraction tools were applied to the HiSoC transcripts to generate text feature sets. The Boruta algorithm was tuned and used to select top features. Factor and regression analyses were performed on top features to identify thematic factors and examine feature associations with UHR, respectively.

Back to article page