.png){kind=logo}
Artificial Intelligence (AI) Measurement Science is an advanced online course offered by Stanford School of Engineering. It provides frameworks and methodologies to evaluate, benchmark, and understand AI systems. The course addresses the growing need for rigorous AI measurement approaches and helps learners connect theoretical principles to real-world use cases, preparing them to critique and implement state-of-the-art evaluation methods.
This course enables learners to:
Evaluation of artificial intelligence systems depends on validity theory which includes content validity, criterion validity, construct validity, external validity and consequential validity.
Research study of psychometric models includes basic item response theory and advanced latent variable models.
examines how scaling laws and intervention methods affect different outcomes which depend on the selected data and computation methods and architectural designs.
Use of synthetic data generation methods to create artificial data which supports its AI evaluation processes.
Examine governance requirements and policy guidelines which affect the evaluation of AI systems.
The course offers a complete online experience through its on-demand format which lasts 10 weeks with an undefined hourly requirement. Students must pay a tuition fee of $4,725 to receive 3 academic credits from Stanford University. The course schedule enables AI research professionals and graduate-level students to study at their own pace.
Validity Theory in AI: Learn frameworks for rigorous AI evaluation
Psychometric Models: Apply item response and latent variable models for AI measurement
Scaling Laws & Interventions: Predict AI performance under different conditions
Synthetic Data for Evaluation: Explore implications and methodologies
Governance & Policy: Understand the societal and ethical dimensions of AI measurement
The candidate needs to hold a bachelor's degree with a minimum GPA of 3.0.
The program serves graduate students and professionals who want to study artificial intelligence through research and evaluation work.
The program suits students who want to pursue careers in artificial intelligence governance and evaluation and advanced AI research.
The course establishes a special solution to the "measurement crisis" in AI through its combination of theoretical foundations and its use of research-based methods. Students acquire practical knowledge through their work with evaluation models and synthetic data generation, which enables them to create authentic and reproducible AI research and assessment.
The program delivers comprehensive training for students who aim to conduct thorough evaluations of AI systems. Theoretical foundations together with methodological approaches and practical applications enable learners to develop skills which will lead to their progress in AI research while they create trustworthy AI systems for ethical use.
