2026
Time Series Forecasting with Irregular Intervals in Applied Behavior Analysis
This study investigates a one-step-ahead forecast of the number of challenging-behavior episodes recorded during therapeutic sessions documented in the SYSABA information system and indicates that explicitly encoding temporal irregularity improves predictive usefulness and that individualized validation is essential for clinically interpretable uncertainty estimates.
J. Kolodziejczyk, Sebastian Limanowski
· International Conference on... · 0 citations