Danna Gurari

This is a small, discussion-based course designed to provide incoming first-year students a foundation to thrive as university scholars, meeting with them from their first day of classes through getting back the results of their first round of midterms. The seminar is a combination of a common curriculum (40%) exploring texts concerning creating an engineering identity, the purpose of an engineering education and the larger values of the college community (mattering, belonging, agency, ownership, inclusivity and service) and a unique curriculum (60%) in which faculty members cultivate these values through their own areas of expertise and interest.

This course covers the fundamentals of neural networks and deep learning as well as how they are used to address many artificial intelligence problems in society. Students will learn to design and implement multi-layered neural network architectures, train them on large amounts of data, and evaluate their performance. Students will also gain practical, hands-on experience by applying learned skills to analyze visual data (computer vision) and textual data (natural language processing).

Computer vision is the subdiscipline of artificial intelligence focused on creating computers that can ‘see’. In this course, students will learn about core and new problems in this field through examination of the types of algorithms commonly used as well as the data employed to train and evaluate those algorithms. The course is taught in a seminar style, with students expected to regularly read and critique research papers from premiere computer vision conferences (such as CVPR, ICCV, and ECCV). As a result, students will also learn how to engage with the latest research in computer vision.

This course will cover core and cutting edge concepts employed in machine learning to solve artificial intelligence problems. Students will learn the theory behind a range of machine learning tools and practice applying the tools to, for example, textual data (natural language processing), visual data (computer vision), and the combination of both textual and visual data.

This course will cover fundamental and state-of-art problems in computer vision, the sub-discipline of artificial intelligence that tries to create computers that can “see”. Students will explore this field through examination of the human-based challenges faced when teaching computers to see.