Link rendezvous states and thruster actions to subsequent motion for data-driven orbital approach modeling and simulation control.

Pair continuous operating telemetry with annotated state changes to evaluate next-stage state prediction for sodium-ion energy storage systems.

A time series dataset covering the overall situation of influenza cases in the United States from 2002 to 2021.

Train models to infer the next step and reachable states in 5G core registration from states, messages, and explicit rules.

The highway visibility in adverse weather estimation image dataset provides high-quality images for improving road visibility assessment under harsh weather conditions.

Connect crack evolution, structural damage, and repair outcome assessment to help models learn verifiable maintenance decisions for concrete structures.

Train underwater glider dynamics models and predict future states with six-degree-of-freedom simulation transitions covering ascent, descent, and turning.

Explore robot state-action-next-state sequences under thruster faults and actuation delays to develop more fault-aware dynamics models.

This dataset is used to predict crowd congestion situations during dismissals of large gatherings.

Trace key events from purchase requisition through payment to support next-event prediction and procurement state-transition modeling.

Connect service health, infrastructure dependencies, and recent events to forecast changes and identify risks earlier.

The ETT time series dataset provides high-quality data support for electric transformer temperature prediction.

Train models to predict the next MOBA lane state from synthetic events and structured pre-event conditions.

The 2020 Weather Indicator Time Series Dataset provides precise weather change data for environmental analysis and predictive models.

Train and evaluate non-cooperative target motion models with dynamics-simulated trajectories and verifiable prediction targets.

Train world models to understand complex underwater interactions with state changes and control feedback across the full docking sequence.

Cross-8 countries daily exchange rate dataset, supporting financial prediction and trading analysis.

Evaluate the stability and state-prediction capability of AR garment tracking with simulated motion, viewpoint changes, state labels, and recovery metrics.

Model and evaluate stream-processing time semantics with simulated watermark, window, and late-event state changes.

Explore simulated state transitions across variations in target spacecraft mass, center of mass, inertia, and contact parameters for prediction and policy training under dynamics uncertainty.