Matrix update: PhD courses re-derived from Proposal-v3

[afarbin]

Teaching-matrix change request — PhD proposed-course columns

Proposal-v3 rewrote the PhD course catalog descriptions; the seven (PhD) columns are re-derived under the standing convention (emphasized → E, named → D, incidental → I).

DATA 5303 (PhD) — was intentionally blank (no description existed); v3 adds one. New cells: version control D, reproducibility D, multithreading/parallel D, Monte Carlo/simulation D, hardware I.
DATA 6301 (PhD) — expanded method list: + regularization, clustering, recommender systems, reinforcement learning, CNNs, optimization D; supervised-vs-unsupervised I→D; model selection stays E; metrics/EDA dropped (no longer named — 6302 carries the applied side).
DATA 6302 (PhD) — now mirrors 6301's method list (practice twin): + logistic regression, trees, SVM, ensembles, regularization, recommenders, RL, CNNs, communication D; model selection/tuning stay E.
DATA 6313 (PhD) — pivot to modern AI: transformers/generative → E (transformers, attention, LLMs, RAG); + transfer learning (fine-tuning) D, metrics D; CNN and applied-vision cells REMOVED (moved to 6301/6302). LLMs, RAG, agentic AI, GNNs recorded as no-taxonomy-row extras.
DATA 5301 (PhD) — cells unchanged (kernel methods & graph-based learning are new but have no taxonomy rows). DATA 5302 / 5390 — descriptions unchanged.

Coarse cells 724 → 743 · granular 1383 → 1403 · pipeline validation: 0 discrepancies.

Derived from PhD/Proposal-v3.docx via the curriculum-review pipeline.

🤖 Generated with Claude Code