Code and results accompanying the paper "Exhaustive Partition Analysis of Speaker-Level Design Effects in Dysarthria Severity Classification." Evaluates a fixed MFCC-based pipeline on the TORGO corpus under leave-one-speaker-out and across all 112 admissible speaker-disjoint partitions, computes per-class recall, confusion matrices, class-balancing controls, and the speaker-level design effect and power bounds reported in the paper. Includes the notebook, generated result files (CSV/JSON), and figures. The TORGO corpus itself is not redistributed and must be obtained from its original distributors.
No takes yet. Share an insight, caveat, or question.
Malik et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: