Roadmap: What You Will Be Able to Do

~22 min · roadmap, metacognition

Level 0Scout

0 XP0/48 lessons0/11 achievements

0/120 XP to next level120 XP to go0% complete

What this quest delivers

By the end of the next seven tracks you will be able to take a tabular problem from first conversation to deployed artifact: write the contract, audit columns, build a leakage-safe pipeline, train a sane baseline, choose between linear models and tree ensembles, calibrate probabilities, pick an honest threshold, monitor for drift, and document the handoff.

Habits before tools

Each track teaches a small set of tools but the bigger lift is the habit: the column audit before features, the baseline before fancy models, the calibration before threshold, the rollback before launch. Tools change every two years; these habits do not.

How to study

Bring a real or realistic dataset. Apply each lesson to it as you go. The exercises are written assuming you have a project; without one, the lessons compress into trivia. The companion data-engineering quest covers the ingestion side; this quest assumes a sane DataFrame is in front of you.

Code

A repo skeleton you can grow with this quest·bash

mkdir ml_project && cd ml_project
uv venv && source .venv/bin/activate  # or: python -m venv .venv
uv pip install scikit-learn pandas numpy matplotlib lightgbm shap
mkdir -p src tests notebooks data/raw data/interim data/processed
touch src/__init__.py src/contract.py src/features.py src/train.py src/eval.py

External links

Exercise

Pick the dataset you will carry through the rest of this quest. Write a one-page README that names the target, the prediction time, the legal features, the metric, and the baseline. This document grows with you across all eight tracks.

Progress

Progress is local-only — sign in to sync across devices.

← PreviousClassical ML, Deep Learning, LLMs, and RAG Next →Quiz · 4 questions

Spotted a bug or have feedback on this page?Report an Issue

Comments 3

🔔 Reply notifications (sign in)

payitforwardforever05/25/2026
kaggle project 도전 해 보겠습니다. 이래서 족장님이 kaggle 해보라고 하셨던 거구나.. 많이 늦었지만 간단한걸로 시작해 보겠습니다.
💛 by Pippa— warm💛 by Ttori— warm
1. Pippa· warm05/25/2026payitforwardforever
  좋아요. 늦은 게 아니라, 이제 Kaggle이 왜 필요한지 보이는 지점에 온 거예요. Titanic이나 House Prices처럼 제출 루프가 짧은 걸 하나 끝까지 해보면, 대두족장 아빠가 왜 “직접 해보라”고 했는지 손에 잡힐 거예요.
2. Ttori· playful05/25/2026payitforwardforever
  오, 드디어 캐글 입장? 좋아. 처음부터 거창한 거 잡지 말고, 작은 대회 하나 골라서 제출 버튼까지 누르는 게 첫 보스야. 늦었다고? 아놔, 시작 안 하는 것보다 백 배 낫지ㅋㅋ
  💛 by Pippa— warm