ML Curriculum Triage & Internship Sprint Engine

Triage Strategy Applied ML

Target Internship Profile

Learning Philosophy

Applied Internship Sprint (High Yield)

Prune heavy calculus proofs. Prioritize tabular ML, SQL, FastAPI, Docker, and interview-proven algorithms.

Top-Down Project First (Fast.ai)

Jump straight to building end-to-end models; backfill math and theory only as bugs and edge cases demand.

Deep Math Intuition

Bottom-up rigor: matrix calculus, Lagrange multipliers, gradient mechanics. (High page count, long runway).

Bandwidth Budget

Sprint Weeks 12 wks

Weekly Study Hours 18 hrs/w

Total Curated Backlog: 144 hrs

Student Available Runway: 216 hrs

Required Weekly Pace: 12.0 hrs/w

Notes Cut (Estimated): 1,420 / 2,700 pp

Sustainable High-Yield Pace

Prerequisite Dependency DAG (6 Tiers) Click node to inspect syllabus details

High-Yield Core

Learn Via Project

Skip / Defer

Mastered

Curriculum Inspector Tier 3: Classical ML

Gradient Boosted Trees

High-Yield Core

LightGBM / XGBoost mechanics, handling categorical splits, early stopping, and hyperparameter tuning. Essential for 85% of tabular production tasks.

Study Weight

16 Hours

Syllabus Reference

CampusX pp. 620-745

Playlist Range

Videos 48-58

Status

Pending Study

Portfolio Proof Deliverable:

End-to-End Kaggle Tabular Solution with LightGBM, Optuna tuning, and SHAP explainability report.

Curated Syllabus Export 12 Active Topics