Unsupervised Machine Learning for Identifying Challenging Behavior Profiles to Explore Cluster-Based Treatment Efficacy in Children With Autism Spectrum Disorder: Retrospective Data Analysis Study

Background Challenging behaviors are prevalent among individuals with autism spectrum disorder; however, research exploring the impact of challenging on treatment response is lacking. Objective The purpose this study was to identify types disorder based engagement in different and evaluate differences between groups. Methods Retrospective data progress for 854 children were analyzed. Participants clustered 8 observed using k means, multiple linear regression performed test interactions skill mastery hours, cluster assignment, gender. Results Seven clusters identified, which demonstrated a single dominant behavior. For some clusters, significant found. Specifically, characterized by low levels stereotypy found have significantly higher than self-injurious behavior aggression (P<.003). Conclusions These findings implications disorder. Self-injurious participants worst response, thus interventions targeting these may be worth prioritizing. Furthermore, use unsupervised machine learning models shows promise.