A new study from researchers at American International University-Bangladesh, Rajshahi University of Engineering and Technology, and Multimedia University in Malaysia developed an auditable, condition-stratified forecasting framework using westbound Interstate I-94 in Minnesota as a single-roadway case study, built around the Metro Interstate Traffic Volume dataset. The researchers began with a data audit, screening hourly records for duplicates, repeated timestamps, physically implausible rainfall observations, and missing hourly intervals; one implausible rainfall event was excluded, and gap-aware 24-hour windowing was applied. The cleaned dataset retained 40,574 usable hourly timestamps, from which 28,855 valid forecasting sequences were derived, each built on a 24-hour input window. The benchmark lineup included a persistence baseline, a 24-hour seasonal-naive reference, LightGBM and XGBoost (the latter tested with and without weather covariates), DLinear, a gated recurrent unit, a compact iTransformer-style model, and the Temporal-Mean Residual Network (TMRNet), with every trainable model run across five prespecified random seeds. Tree-based models led: XGBoost without weather variables achieved the lowest mean test mean absolute error, 141.67 ± 0.31 vehicles per hour, followed by LightGBM at 142.63 ± 0.41 and all-feature XGBoost at 142.89 ± 0.56, while TMRNet trailed at