Fase 5 — Machine Learning Pemula

Estimasi: 1.5 minggu (~30 jam dari kelas Dicoding 90 jam — pre-cover sepertiganya) Prasyarat: Fase 1-4 selesai Tujuan: Paham fondasi ML klasik + sklearn API. Saat kelas Dicoding ML 90 jam dimulai, kamu sudah ahead.

Strategi

Kelas Dicoding "Belajar Machine Learning untuk Pemula" akan 90 jam detail. Kita TIDAK mau habiskan semua sekarang — biarkan kelas yang dalam.

Yang kita lakukan: kuasai konsep inti + sklearn API + 1 end-to-end project. Sehingga saat kelas dimulai, kamu sudah punya mental model dan bisa fokus pada:

• Aplikasi advanced
• Deeper algorithms
• Capstone project quality

Roadmap

File	Topik	Estimasi
01-ml-workflow.md	End-to-end ML pipeline	4 jam
02-supervised-classification.md	Logistic, Tree, RF, KNN, SVM	5 jam
03-supervised-regression.md	Linear, Ridge, Lasso, Tree-based	4 jam
04-unsupervised.md	K-Means, PCA, DBSCAN	4 jam
05-feature-engineering.md	Encoding, scaling, selection	4 jam
06-evaluation.md	Metrics, CV, ROC, hyperparameter tuning	4 jam
07-kaggle-project.md	End-to-end Kaggle submission	5 jam
challenges.md	Final challenges	-

Tools

pip install scikit-learn xgboost lightgbm

Aturan

1. Kerjakan di Jupyter Notebook
2. Setiap algoritma harus train minimal 1 kali pada data nyata
3. Submit ke Kaggle (skor tidak penting, prosesnya yang dihargai)

---

📚 Referensi Belajar Fase 5

⭐ Wajib

Resource	Tipe	Harga
StatQuest — Machine Learning Playlist	🎥 Video (50+ ep)	🆓
Scikit-learn User Guide	📖 Teks	🆓
Kaggle Learn — Intro to ML	💻 Interaktif (3 jam)	🆓

Recommended

Resource	Tipe	Harga
Andrew Ng — ML Specialization	🎥 Course	🆓 audit / 💰 $49/bln
Kaggle Learn — Intermediate ML	💻 Interaktif	🆓
Machine Learning Mastery	📖 Blog	🆓
Kaggle Competitions — Getting Started	💻 Kompetisi	🆓
"Hands-On Machine Learning" — Géron	📚 Buku	💰 ~$60
"Introduction to Statistical Learning" (ISLR)	📚 Buku	🆓 PDF

Lihat daftar lengkap di RESOURCES.md

---

Mulai dari: 01-ml-workflow.md