Data Science & Artificial Intelligence
Master Hub & Knowledge Base
A complete 16-session (8-week) learning path aligned with Indonesia’s National Work Competency Standard (SKKNI No. 299/2020). Enriched with mathematical formulations, architecture diagrams, and modern applied Python code.
16-Session Learning Curriculum
Each session is built in depth with industry research, theoretical formulations, and real Python implementations.
Note: English translations are in progress. Sessions not yet translated open the original Indonesian module.
AI Foundations & Computing Environment
Set up a modern Python 3.10+ ecosystem, Jupyter/Colab, GPU acceleration, the four AI paradigms, PEAS agents, and Indonesia’s AI National Strategy 2020–2045.
Methodology & Data Science Tools
CRISP-DM lifecycle, 21 SKKNI 299/2020 units, Git/GitHub, the 6 core libraries, NumPy vectorization, and a real immunization dataset case study.
Business & Data Understanding
Asymmetric financial costs, end-to-end machine learning, Tata Global stock prediction, Stevens’ four scales, and automotive data analysis.
Visualization & Data Cleaning
Variable vs. statistical visualization, one-way ANOVA, five statistics notebooks, scientific sampling, SMOTE imbalance handling, and real messy-data cleaning.
Data Construction & Feature Engineering
Numeric scaling (MinMax/Robust/Standard), distribution transforms (Box-Cox), categorical encoding, data integration (merge/join), PCA, and feature selection.
Supervised Learning
Leakage-proof validation scenarios, classification algorithms (SVM, Decision Tree, Naive Bayes, KNN, AdaBoost) and regression (OLS, SVR, Random Forest).
Clustering & Neural Networks
Unsupervised clustering (K-Means++, Hierarchical, DBSCAN, Silhouette) and deep learning architectures (Perceptron, MLP, CNN, RNN, Backpropagation).
Model Evaluation & MLOps Deployment
Full evaluation metrics (ROC-AUC, PR-Curve, SHAP), model serialization, web deployment (Flask/Heroku, FastAPI/Streamlit), and model drift monitoring.