01
Data Engineering Foundations
พื้นฐานวิศวกรรมข้อมูล
134 slidesPowerPoint · 5.18 MB
Download ↓DATA SCIENCE × DATA ENGINEERING
See the data, the system and the decision together.
This hub orders lessons from data foundations through features and machine learning to production. You need not open every page—but should know where each lesson sits in the whole system.
เส้นทางการเรียน01—0530+ lessons · 5 slide collections · live laboratories
DATA FOUNDATION · PATH 01
Begin with operational systems and data meaning
Understand transactions, shared data contracts, and why a logical view is a beginning rather than the final answer.
01
VIEW & DATA CONTRACT
02Build a stable, reusable and testable data interface.
Open lesson →BULK UPDATE vs REBUILD
03Understand when mass UPDATE costs more than read–transform–new table.
Open lesson →OLTP → FEATURE
04Trace transactions, CDC and history into point-in-time features.
Open lesson →DATA SYSTEMS LAB
Compare PostgreSQL, DuckDB, ClickHouse, materialization and IVM.
Open lesson →PIPELINES & ANALYTICAL STORAGE · PATH 02
Make movement and repeated computation trustworthy
Move from orchestration to OLAP and feature storage while separating freshness, physical work and operating cost.
01
AIRFLOW
02Run DAGs while inspecting dependencies, retries, schedules and TaskFlow.
Open lesson →RELIABLE WEB SCRAPING
03Experiment with selectors, schema drift, missing data, duplicate keys and HTTP 429 in a safe simulated site.
Open lesson →OLAP → FEATURE STORAGE
04Compute from history and materialize only valuable reusable state.
Open lesson →FEATURE STORE
Version features and control point-in-time correctness and serving parity.
Open lesson →FEATURE ENGINEERING · PATH 03
A feature is costly, contextual evidence—not merely a formula
Study shared computation, representation and domain meaning before features reach a model.
01
COMPOSITE & REUSE
02Find shared structure and avoid repeated computation.
Open lesson →RFM & CUSTOMER MOVEMENT
03Select a customer, calculate RFM and read slopes as churn signals.
Open lesson →REPRESENTATION
04PCA, SVD, projections, embeddings and model-derived scores with leakage control.
Open lesson →BIOINFORMATICS
05Trace complexity from sequence scans to structure and embeddings.
Open lesson →FINANCIAL DATA
06Interpret trend, momentum, volatility, liquidity and risk carefully.
Open lesson →12-DOMAIN ATLAS
07Compare data, formulas, compute time and meaning across twelve domains.
Open lesson →DEEP LEARNING
See what representation learning automates—and cannot recover.
Open lesson →SQL & MACHINE LEARNING · PATH 04
Begin with the formula, then choose the right abstraction
Build models from relational operations before moving to declarative interfaces, libraries and recommenders.
01
FORMULA → MARIADB SQL
02Derive kNN, Naive Bayes, regression, k-means and Z-score.
Open lesson →PURE SQL ML
03Open six ML families without Python.
Open lesson →SQL RECOMMENDER
04Build candidates, similarity and ranking in SQL.
Open lesson →DECLARATIVE ML
05Separate SQL-native, SQL-callable and commercial platforms.
Open lesson →MADLIB LAB
Use an in-database library after understanding what it hides.
Open lesson →PRODUCTION & ONLINE ML · PATH 05
Close the loop from features to prediction, drift and continuous learning
Separate online inference, online feature computation and true online learning while preserving evidence for delayed labels.
01
ONLINE ML · 3 PATTERNS
02Understand three production meanings of online.
Open lesson →DUCKDB × CLICKHOUSE
03Compare embedded and server OLAP execution boundaries.
Open lesson →ML DRIFT · CHAMPION/CHALLENGER
04Simulate monthly electricity data, inspect feature drift, compare algorithms and training windows, then decide whether to retain or replace the Champion.
Open lesson →TEACHING LAB INDEX
Continue to cross-course labs in algorithms, bioinformatics and teaching design.
Open lesson →DOWNLOAD LECTURE MATERIALS
Five slide collections and hands-on files
The PowerPoints preserve diagrams and lecture flow; the notebooks and labs provide hands-on practice.
02
Data Storage and Data Systems
การจัดเก็บข้อมูลและระบบข้อมูล
65 slidesPowerPoint · 1.72 MB
Download ↓02B
XPath, BeautifulSoup and Selenium
Screen Scraping ตั้งแต่ Selector ถึง PostgreSQL
36 slidesPowerPoint · 0.13 MB
Download ↓02C
Scraping Notebooks and Data
Notebook และข้อมูลสำหรับฝึก Scraping
2 notebooksZIP · 0.06 MB
Download ↓03
Moving and Processing Data at Scale
การเคลื่อนย้ายและประมวลผลข้อมูลขนาดใหญ่
19 slidesPowerPoint · 0.38 MB
Download ↓04
Beyond Data Engineering
ต่อยอดจากวิศวกรรมข้อมูล
20 slidesPowerPoint · 0.35 MB
Download ↓COURSE PRINCIPLE
Data science asks what data means. Data engineering makes the answer reproducible.
A strong model cannot repair mistimed data, leaky features or an unexplainable pipeline. The goal is not only prediction, but knowing which system produced the answer and how far it can be trusted.