Course Syllabus: Trustworthy Machine Learning
Course Information
Course Title: Trustworthy Machine Learning
Institution: UCLA Extension
Semester: Fall 2025
Format: Online Only
Duration: 11 weeks
Course Link: COM SCI X 450.44
Prerequisites: Machine Learning fundamentals, Linear Algebra, Statistics, Python programming
Credits: 3
Course Description
This course provides a comprehensive introduction to the principles, methods, and applications of trustworthy machine learning. Students will learn to design, implement, and evaluate ML systems that are fair, robust, transparent, privacy-preserving, and accountable. The course combines theoretical foundations with hands-on projects using real-world datasets and modern tools.
Learning Objectives
By the end of this course, students will be able to:
- Understand Core Concepts: Define and explain the key principles of trustworthy ML including fairness, robustness, interpretability, privacy, and accountability
- Identify Vulnerabilities: Recognize potential sources of bias, adversarial attacks, and privacy leaks in ML systems
- Apply Mitigation Techniques: Implement state-of-the-art methods for bias mitigation, adversarial defense, and privacy protection
- Evaluate Systems: Use appropriate metrics and evaluation frameworks to assess the trustworthiness of ML models
- Design Solutions: Architect end-to-end trustworthy ML systems for real-world applications
Course Topics
| Week | Topic |
|---|---|
| 1 | Introduction to Trustworthy ML |
| 2 | Model Evaluation and Fairness |
| 3 | Midterm Project |
| 4 | Privacy Enhancing Technologies I |
| 5 | Privacy-Enhancing Technologies II & Federated Learning |
| 6 | Gen AI Security Models and Frameworks |
| 7 | Safety, Alignment, and Evaluation in LLMs |
| 8 | Advanced LLM Safety |
| 9 | Security Testing and Red Teaming |
| 10 | AI Regulatory Frameworks and Compliance |
| 11 | Final Project |
Required Resources
Software & Tools
- Python 3.8+ with scikit-learn, PyTorch/TensorFlow
- Fairness toolkits: AIF360, Fairlearn
- Privacy libraries: Opacus, PySyft
- Interpretability tools: SHAP, LIME, Captum
Computing Resources
- Google Colab Pro or similar cloud platform
This syllabus is subject to change with advance notice. Check the course website regularly for updates.