Uncertainty Modeling in AI | Lecture 12 (Part 2): Graph cut and alpha expansion

Here's the video lectures of CS5340 - Uncertainty Modeling in AI (Probabilistic Graphical Modeling) taught at the Department of Computer Science, National University of Singapore (NUS).

This is an introductory course on probabilistic graphical modeling which was recorded for online learning at NUS due to COVID-19. The topics covered include:

Lecture 1: Introduction to probabilistic reasoning
Lecture 2: Bayesian networks (Directed graphical models)
Lecture 3: Markov random fields (Undirected graphical models)
Lecture 4: Variable elimination and belief propagation
Lecture 5: Factor graph and the junction tree algorithm
Lecture 6: Parameter learning with complete data
Lecture 7: Mixture models and the EM algorithm
Lecture 8: Hidden Markov Models (HMM)
Lecture 9: Monte Carlo inference (Sampling)
Lecture 10: Variational inference
Lecture 11: Variational Auto Encoder and Mixture Density Networks
Lecture 12: Graph cut and alpha expansion

Disclaimer: This video lecture is provided freely for your reference. The lecturer and NUS are not responsible for anything expressed herein.

[qiqinzhan]

Hi professor, I have a question in 45:50 why the unary cost on the upper left edge that associate a and (not alpha) is not U(not alpha), instead it is marked as U(alpha)?