Tracking the PhD journey
Self-assessment as of February 16, 2026. Later milestones below reflect repository activity through July 2026
Systematic review of 138 papers, identified research gap
Three-layer hybrid architecture designed and documented
Successfully replicated VenSim model in MATLAB (R² > 0.99)
Validated PCA/DMD analysis pipeline on benchmark system
First full hybrid model implementation
11-figure IEEE-formatted defense package prepared
Status-update slide deck and talking points for the committee
EORTC preprint + pre-registered data request finalized; standalone benchmark vs SD-only / Q-PRIMA arms; methods paper draft v6; network size grounded in Cloyes 2022 (per-run draw)
EORTC AYA QLQ-C30 data request → held-out predictive validation
Dissertation defense, revisions, submission
Problem statement, research questions, dissertation structure
138 papers, systematic review, gap analysis
Three-layer architecture, mathematical formulation
Method validation on benchmark system
Full hybrid model implementation
Ablation ladder, benchmark docking, scenario analysis
Interpretation, limitations, future work
Summary, contributions, future directions
The support-network mean degree is now drawn per run from the empirical AYA distribution (Cloyes 2022: mean 6, range 2-10) instead of a fixed value. Headline results unchanged; between-patient network heterogeneity now emerges. Methods paper updated to draft v6; all figures and tables regenerated.
Finalized the preprint (verified architecture, pattern-oriented results, pre-registered external-validation data request) and the standalone benchmark comparing the full hybrid against SD-only and Q-PRIMA arms: the full coupled model reaches 7.0% validation error across seeds 1 to 20.
Prepared the model status-update deck and talking points covering the coupling formalism, verification results, and the external-validation plan.
Created MATLAB script package to generate all 11 IEEE-formatted figures for the 30-minute committee presentation. Includes architecture diagrams, PCA results, and scenario comparisons.
Verified all 138 references. Found 2 hallucinations and 1 incorrect citation (wrong journal). Created Excel workbook with 3-pass verification methodology.
Translated the Q-PRIMA model of Hasgul from VenSim to MATLAB. Achieved R² > 0.99 correlation with the original trajectories. The result validates the implementation pipeline.