We are pleased to announce a new course, CIS 700: Hybrid Artificial Intelligence that will be offered this fall at Syracuse University. The course description can be found below along with a “test lecture” on YouTube.
Today’s state-of-the-art machine learning systems such as deep neural networks have shown impressive capabilities for perceiving and generating information by learning models from data but often fail at reasoning tasks. Symbolic systems, on the other hand, excel in various forms of inference but often struggle with learning. Recent advances in systems that combine machine learning with symbolic artificial intelligence present promising new directions toward the combination of reasoning and learning. In this course, we examine fundamental concepts related to hybrid artificial intelligence as well as several major lines of research in this area. The course will start with initial lectures on dual process theory, symbol grounding, and metacognition. The remainder of the course will include the following topics: neurosymbolic AI (e.g., logic tensor networks, logical neural networks), abductive learning (to include general discussions on abduction), inductive logic programming, cognitive inspired systems (to include hyperdimensional computing), symbolic regression, and related topics.