New Spring 2020 Class Offering Training in AI to All Duke Students
Class Title: "AI for Everyone"
EGR 190.06 (undergraduate students) / EGR 590.06 (graduate students)
This class will introduce the student to machine learning (ML) and artificial intelligence (AI) methods that have become increasingly useful in practice, specifically deep learning and neural networks. Application areas include image analysis, text analysis, and optimal decision making. This class is directed to any Duke student, independent of major, who is interested in learning the basics of ML and AI. The class will be presented in a manner that places primary emphasis on the intuition behind the methods, with the mathematical details kept at as simple a level as possible. While the mathematical detail will be kept as simple as possible, the ability to code in software (Python) will be important. The students will be introduced to the coding software behind AI in practice. While prior experience in software is not a prerequisite, it is a plus. The student should be interested in learning to code AI models, and should be prepared to learn how to do so.
The class content will be delivered in several complementary ways:
1. Students will complete the Duke-developed machine learning material on Coursera:
2. The Coursera content will be partitioned into weekly assignments, and a weekly in-person review session will be held at an assigned time. These weekly in-person sessions will complement the aforementioned Coursera content, providing review of key concepts and examples of how the methods are applied in practice. This will manifest a so-called flipped-class teaching framework, with detailed technical instruction offered via the aforementioned (free) online content, and the in-person sessions will be devoted to review, answering student questions, and demonstration of applications of the methods.
3. The student must complete at least four in-person-learning experiences (IPLEs) offered within the Duke +DS program (https://plus.datascience.duke.edu/learn-ds). The IPLEs cover various aspects of machine learning, implementation details, applications, and the ethics of machine learning. The variety of IPLE topics presented over a semester are meant to constitute a diverse set of content, meant to cover a wide range of student interests. Within the IPLEs, material will be presented for students interested in engineering, natural science, and humanities applications of AI.
4. Each student will complete a project, in which machine learning will be applied to a problem of interest to the student. Suggested projects will be provided, in a wide range of areas of studies (in the natural sciences, social sciences, humanities and engineering).
This class is for 0.5 credits (undergraduate students) or 1.5 credits (graduate students), and students will get a regular grade. There will be a weekly meeting of the students, in Gross Hall 330, Mondays 4:40-6:10pm. The class is open to all Duke students.