The Dynamic Latent-State Model treats student engagement in self-paced learning as a continuous, multi-dimensional latent process — inferred from noisy observations, tracked with particle filtering, and estimated with PMCMC.
Xₜ = ( Sₜ , Eₜ , Pₜ , Rₜ )ᵀ
The tutorial walks you through every step with runnable code. The user guide is the complete reference for research and deployment.
Learn by doing
A progressive, six-chapter walkthrough. Every code block runs. You will simulate, filter, estimate, and deploy the model by the end.
The complete reference
Everything you need to use the library in research or production: API, workflows, diagnostics, performance, and best practices.
From engagement theory to stochastic calculus to Bayesian computation — the DLSM sits at an intersection.