A Reproducible Statistical-Learning Framework for Sparse Longitudinal Transaction Data: Multiplicity-Screened Association Rules, Temporal Validation, and Generalized Linear Models

Sparse longitudinal transaction data create a statistical-learning problem in which low basket occupancy, combinatorial candidate growth, temporal heterogeneity, and data-quality sensitivity complicate interpretable pattern discovery. This study proposes a reproducible, sparsity-aware framework that separates global distributional characterization, unsupervised co-occurrence learning, statistical qualification, adjusted outcome modeling, and transparent prioritization. The empirical application comprises 31,157 baskets, 139,396 valid item lines, 710 products, and 762 active days; the 31,157 × 710 basket-product representation has 0.630% density. Multiplicity-screened association-rule learning retains 1,174 directional rules. In a strictly subsequent validation period, 93.6% of discovery rules preserve lift above one, 49.3% satisfy all prespecified thresholds again, and rank correlations for support, confidence, and lift range from 0.803 to 0.849. An estimated-dispersion NB2 model with day-clustered uncertainty preserves the principal basket-breadth associations, while Gamma sensitivity analysis confirms the direction, but not invariant magnitude, of late-service basket-value effects. The results show that sparse transactional structure is studied more defensibly when algorithmic discovery is explicitly separated from multiplicity screening, statistical inference, temporal replication, and decision prioritization. Because the empirical evaluation is a longitudinal single case, the reported effect magnitudes remain case-specific; the transferable contribution is the auditable statistical-learning procedure.