Team — Pakistan.AI

Neural Networks course project · National Yunlin University of Science and Technology

Group: Pakistan.AI · Course: Neural Networks · Instructor: Prof. Hsuan-Ting Chang · Affiliation: FIC Lab, National Yunlin University of Science and Technology

Members

Member Student ID Profile
Raja, Muhammad Junaid Ali Asif M11217073 ORCID 0009-0008-9249-9983 · FIC Lab, National Yunlin University of Science and Technology
Sultan, Adil M11217078 FIC Lab, National Yunlin University of Science and Technology
Hassan, Shahzaib Ahmed M11217081 FIC Lab, National Yunlin University of Science and Technology

All three members contributed to the design, experiments, analysis, and this report.

Contributions

  • Validation spine & metric (src/cv.py), the frozen patient-grouped folds, and the cv-guardian no-leak discipline.
  • Tabular feature engineering & GBDT (src/features.py, src/gbdt.py) — the efficient anchor, including the patient-relative ugly-duckling features and the underfit-bug fix.
  • Image experts & the 12-backbone sweep (src/vision/*) — the primary backbone and the resolution/regularizer ablations.
  • Efficiency frontier & stacking (src/efficiency.py, src/stack.py, reports/frontier.py) — the three-axis cost measurement and the trivial-combiner result.
  • EDA, performance analysis, and this companion site — the figures, the negative-results catalogue, and the write-up.

Citing

To cite the repository, use the metadata in CITATION.cff (GitHub renders a “Cite this repository” button from it). Please also cite the SLICE-3D dataset and the official metric — see References.