CS 257: Logic and Artificial Intelligence (PHIL 356C)
This is a course at the intersection of philosophical logic and artificial intelligence. After reviewing recent work in AI that has leveraged ideas from logic, we will slow down and study in more detail various components of high-level intelligence and the tools that have been designed to capture those components. Specific areas will include: reasoning about belief and action, causality and counterfactuals, legal and normative reasoning, natural language inference, and Turing-complete logical formalisms including (probabilistic) logic programming and lambda calculus. Our main concern will be understanding the logical tools themselves, including their formal properties and how they relate to other tools such as probability and statistics. At the end, students should expect to have learned a lot more about logic, and also to have a sense for how logic has been and can be used in AI applications. Prerequisites: A background in logic, at least at the level of
Phil 151, will be expected. In case a student is willing to put in the extra work to catch up, it may be possible to take the course with background equivalent to
Phil 150 or
CS 157. A background in AI, at the level of
CS 221, would also be very helpful and will at times be expected. 2 unit option only for PhD students past the second year. Course website:
http://web.stanford.edu/class/cs257/
Terms: Win
| Units: 2-4
Instructors:
Icard, T. (PI)
;
Zaffora Blando, F. (TA)
PHIL 356C: Logic and Artificial Intelligence (CS 257)
This is a course at the intersection of philosophical logic and artificial intelligence. After reviewing recent work in AI that has leveraged ideas from logic, we will slow down and study in more detail various components of high-level intelligence and the tools that have been designed to capture those components. Specific areas will include: reasoning about belief and action, causality and counterfactuals, legal and normative reasoning, natural language inference, and Turing-complete logical formalisms including (probabilistic) logic programming and lambda calculus. Our main concern will be understanding the logical tools themselves, including their formal properties and how they relate to other tools such as probability and statistics. At the end, students should expect to have learned a lot more about logic, and also to have a sense for how logic has been and can be used in AI applications. Prerequisites: A background in logic, at least at the level of
Phil 151, will be expected. In case a student is willing to put in the extra work to catch up, it may be possible to take the course with background equivalent to
Phil 150 or
CS 157. A background in AI, at the level of
CS 221, would also be very helpful and will at times be expected. 2 unit option only for PhD students past the second year. Course website:
http://web.stanford.edu/class/cs257/
Terms: Win
| Units: 2-4
Instructors:
Icard, T. (PI)
;
Zaffora Blando, F. (TA)
Filter Results: