Identifying latent subgroups of children with developmental delay using Bayesian sequential updating and Dirichlet process mixture modelling.

ABSTRACT: Identifying children who are at-risk for developmental delay, so that these children can have access to interventions as early as possible, is an important and challenging problem in developmental research. This research aimed to identify latent subgroups of children with developmental delay, by modelling and clustering developmental milestones. The main objectives were to (a) create a developmental profile for each child by modelling milestone achievements, from birth to three years of age, across multiple domains of development, and (b) cluster the profiles to identify groups of children who show similar deviations from typical development. The ensemble methodology used in this research consisted of three components: (1) Bayesian sequential updating was used to model the achievement of m

SUBMITTER: Gilholm P 

PROVIDER: S-EPMC7266333 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature