Existing literature reveals that overall noise levels in pediatric and neonatal hospital units are exceeding acceptable ranges and exhibiting persistent noise level fluctuations over time and location. Previous studies have also revealed insights into the subjective perception of noise and explored existing noise sources and their measured acoustical properties. However, less has been done regarding clear identification of hospital noise sources related to staff annoyance in addition to detailed, informative representations of the acoustical characteristics of hospital environments. This study utilizes two methods of unsupervised learning techniques— factor analysis and clustering analysis— to assess occupant perception alongside detailed noise level measurements. Data were collected in two pediatric and neonatal hospital units to provide informative representations of the existing acoustical environments. The factor analysis results show three inherent noise categories among the various noise sources in ...