Practical AI and Data Handbook for Engineers

Practical AI and Data Handbook for Engineers

A text-based course for engineers who can verify numbers produced by AI, instruct AI accurately, and explain its conclusions

• Structure: 8 sections, 25 lessons, 80 practice questions, 4 hands-on notebooks, and 1 handbook PDF

Recommended for people like these

• Engineers who use AI-generated analyses and reports in their work and need criteria to independently verify whether the numbers are correct

• A data professional who wants to know what to ask again when told that a report is 99% accurate

• Quality and process professionals who want to reorganize control charts, process capability, and measurement system analysis by connecting them with AI analysis results

What You Will Learn

Section Contents
1. Principles of AI and LLMs and safe usage: tokens, context, temperature, hallucinations, RAG and fine-tuning, information leakage, and prompt injection
2. AI output verification and collaborative instruction design: types of hallucinations and the verification ladder, invariant checks, and report structure
3. Probability and statistics: intuition for the normal distribution, standard error and p-values, base rates, and Simpson’s paradox
4. Machine learning evaluation metrics and overfitting: validation design, recall, precision, F1, MCC, regression metrics, and imbalanced data
5. Data quality and leakage: data checklists, five types of leakage and splitting principles, drift, and PSI
6. SPC and process capability: control charts and rules for identifying anomalies, the difference between control limits and specification limits, Cp, Cpk, and sigma levels
7. Measurement System Analysis (MSA): bias, linearity, and stability, %GRR and ndc in Gage R&R, and the kappa coefficient
8. AI cases in manufacturing: predictive maintenance, visual inspection, virtual metrology, yield models, and an AI adoption checklist

Provided Materials

• 25 class notes (ready to read in the classroom), 25 accompanying class materials PDFs—one for each class—and a combined PDF compiling all 25 classes

• Handbook PDF, *Practical AI and Data Handbook for Engineers*, 48 pages

• An 80-question section-by-section practice workbook and a PDF of answers and explanations.

• Section-by-section web mock exam: Solve 10 practice questions online and immediately view your score and explanations. During the exam, you can ask the AI assistant up to 3 questions per attempt.

• 4 practice notebooks (control charts, Cp·Cpk, Gage R&R, and imbalanced classification metrics). Download them from notebooks.zip in the course materials.