Spatio-temporal parse network-based trajectory modeling on the dynamics of criminal justice system

Area of Science:

• * Statistical modeling
• * Spatio-temporal analysis
• * Network analysis

Background:

• * Existing group-based trajectory modeling has limitations in capturing complex, evolving structures.
• * Real-world problems often involve dynamic neighborhood structures that are challenging to model.
• * There is a need for advanced frameworks to analyze nonlinear trajectories.

Purpose of the Study:

• * To propose a novel network-based trajectory modeling framework.
• * To represent evolving neighborhood structures using a spatio-temporal parse network (STPN).
• * To develop a hierarchical model integrating latent field representation with STPN.

Main Methods:

• * Design and analysis of a spatio-temporal parse network (STPN).
• * Development of a hierarchical model with latent field representation.
• * Application of spatial random effects to model heterogeneity and autocorrelation.
• * Investigation of trajectories under spatial and temporal dependence structures.

Main Results:

• * The proposed STPN framework provides a principled qualitative and quantitative specification for trajectory modeling.
• * The hierarchical model effectively merges latent field representations with the STPN.
• * Spatial random effects successfully characterize location-based heterogeneity and autocorrelation.
• * The framework allows for accessible diagnosis and modeling of complex developmental trajectories.

Conclusions:

• * Network-based trajectory modeling using STPN offers a powerful approach for analyzing nonlinear dynamics.
• * The developed framework simplifies the understanding and modeling of complex spatio-temporal processes.
• * This methodology is applicable to diverse fields, including criminal justice empirical processes.