Events

Toward Interpretable and Sustainable AI/ML

Speaker: C.-C. Jay Kuo

Location: 370 Jay Street, Room 825
Videoconference link: https://nyu.zoom.us/j/94677623510

Date: Tuesday, April 23, 2024

Rapid advances in artificial intelligence (AI) and machine learning (ML) have been attributed to the wide applications of deep learning (DL) technologies. There are however concerns with this AI wave. DL solutions are a black box (i.e., not interpretable) and vulnerable to adversarial attacks (i.e., unreliable). Besides, the high carbon footprint yielded by large DL networks is a threat to our environment (i.e., not sustainable). It is important to find alternative AI technologies that are interpretable and sustainable. To this end, I have conducted research on green AI/ML since 2015. Low carbon footprints, small model sizes, low computational complexity, and mathematical transparency characterize green AI/ML models. They differ completely from DL models since they have neither computational neurons nor network architectures. They are trained efficiently using labels (but no backpropagation). Green AI/ML models offer energy-effective solutions in cloud centers and mobile/edge devices. They consist of three main modules: 1) unsupervised representation learning, 2) supervised feature learning, and 3) decision learning. Green AI/ML has been successfully applied to various applications. I will use several examples to demonstrate their effectiveness and efficiency.