Syllabus
ML for Innovation Research: PhD/DBA Seminar — Prof. J. Osterrieder
Overview
| Format | 2 × 3-hour hands-on sessions |
|---|---|
| Level | PhD / DBA |
| Prerequisites | Basic Python, introductory statistics |
| Dataset | Swiss Innovation Survey (500 synthetic projects) |
Session 1: Unsupervised ML (3 hours)
Theme: Discovering patterns and structure in innovation data
| Module | Topic | Techniques |
|---|---|---|
| 01 | The Map | ML paradigms, workflow, features |
| 02 | The Discovery Branch | K-Means, PCA, UMAP |
| 03 | The Text Branch | VADER, TF-IDF, LDA |
| 04 | Where Branches Cross | Cross-tabulation, multi-method design |
Session 2: Supervised ML & GenAI (3 hours)
Theme: Predicting outcomes and leveraging AI for research
| Module | Topic | Techniques |
|---|---|---|
| 05 | The Prediction Branch | Features, labels, feature importance |
| 06 | The Prediction Branch | Random Forest, Logistic Regression, ROC/AUC |
| 07 | Beyond the Tree | LLMs, prompt engineering, structured output |
| 08 | The Complete Map | Decision framework, thesis patterns |
Learning Outcomes
- Apply unsupervised and supervised ML to innovation research data
- Process and analyze text data using NLP techniques
- Evaluate and report ML model results for academic publications
- Use generative AI responsibly as a research tool
- Design multi-method ML research approaches for DBA/PhD theses