AI and Machine Learning on Graphs
You know artificial intelligence (AI) is good at distinguishing cats from dogs. You also know machine learning is all that makes Siri work. However, those familiar tasks are just the tip of the iceberg. Most data are not represented so well as digital images or natural language, and there is no way to apply popular AI methods such as the abused “CNN." For complex structured data, especially relational data, a “graph” is a natural model. A graph is anything that contains nodes and edges. For instance, we can consider the DKU community as a big “graph” according to WeChat connections. You and I, and everyone else is a “node." If I am your WeChat contact, then we are connected by an “edge.” In this course, we will study popular approaches to dealing with graph data in AI and machine learning (such as how Google gives you search results and Douban gives you movie recommendations). We will develop mathematical fundamentals of graph data using basic linear algebra and use those to process graph data so that they can be plugged into a machine learning framework. We will then work with interesting datasets and see the power of AI and machine learning for graphs!
Prerequisites: Linear Algebra, Basics of Machine Learning, and familiarity with Python
Instructor: Dongmian Zou
Basic Digital Humanities: Using Databases in Historical Research
You have a great Signature Work idea but don’t have enough case studies to test it? You want to use some visual materials to illustrate your project but have no idea where to get them? You suspect that something (a book, an idea, a commodity, for example) was popular in a particular historical time but feel impossible to verify the suspect? Your intuition tells you that the historical figures A and B must have had some direct interactions but don’t know how to find evidence? What you need is database. Digital humanities are revolutionizing how we study history, and the use of databases of digitalized materials — newspapers, magazines, pictures, government files, etc. — is a basic but probably the most useful method in this endeavor. This course leads you to enter this new world. After quickly reviewing today’s digital humanities and introducing some inspiring examples, the course will be customized according to your specific project. It will help you identify databases that are suitable for your project and experiment what you can do with them. By the end of the course, you will not only collect enough materials to fuel your Signature Work, these materials will push you to think about your project more critically and innovatively.
Instructor: Xianjun Feng
Climate Risk, Institutional Investors, and Corporate Finance
Business investment in sustainable and socially responsible projects has become a mainstream voice among all investors. Institutional investors play a pivotal role in committing to environmental, social, and corporate governance (ESG) principles as part of their investment portfolios and strategies. The nascent empirical literature in corporate finance employs a series of causal inference techniques to identify the causal relationship between institutional investors and corporate performance in environmental and sustainable development. Open but important research topics include whether institutional investors consider climate risks in their investment portfolio, whether institutional investors drive firms’ corporate social responsibility or ESG performance, and how climate risks would affect investors’ returns. This mini-term course provides some quantitative and causal evidence on these and related questions. It presents cutting-edge interdisciplinary research articles in the fields of environmental economics and corporate finance. Students will develop a better understanding of how institutional investors respond to climate risk and corporate CSR or ESG performance and learn the state-of-the art causal inference methods applied in the field of empirical corporate finance. They will navigate research interests and propose a research question for a Signature Work project.
Prerequisites: Econ 201 Intermediate Microeconomics I and Econ 203 Introduction to Econometrics
Instructor: Cui Jingbo
Doing Ethnographic Fieldwork for Signature Work
This mini-term course will guide students step-by-step through the process of doing ethnographic fieldwork. It is designed for students who are considering or using fieldwork as a research method for data collection in their Signature Work. Students will develop an in-depth understanding of ethnographic fieldwork as a method, its advantages and limitations, and a firsthand experience of doing it through a mock project – either a prospective project, a sub-project or an actual Signature Work project that is ongoing. First, each student will propose a project. Then the student will refine their proposed project by learning how to design a project of the right scope and ask the right kinds of research questions through examples and discussion. Second, each student will design a research plan for the proposed project. Third, we will collect and analyze fieldwork data. Lastly, we will produce a short research paper based on the data collected. The class will take the forms of 1) active discussions, involving reading and commenting on each other’s work, such as a “mock proposal defense," and 2) hands-on exercises such as semi-structured and informal interviews, participant observation, and taking and organizing notes.
Instructor: Keping Wu
Fintech and AI: Research and Practice
Bitcoin! Blockchain! Alpha Go! Smart contract! Do you know financial technology (fintech) and artificial intelligence (AI), once mysterious, are changing the world economy dramatically? In this course, we will overview the frontier of fintech and AI in both research and practice. Students will learn in experiential education the three research methodologies for this topic: Reinforcement learning, behavioral experiments, and interdisciplinary big data. Moreover, students will explore in business cases how fintech and AI can empower decentralized finance applications. At the end of this course, students learn collaboratively in teams that include economists, data scientists and data engineers to propose a project of both intellectual merit and practical impact. No mandatory prerequisite. This course is open to a diverse background of students who are interested in contributing to advancing the frontiers of fintech and AI. The course is especially relevant for third-year students with a background in mathematics, economics, data science, computer science, or related fields.
Instructor: Luyao Zhang
Invitation to Spacetime
This course will introduce students to the mathematical theory of spacetime which defines our fundamental picture of the universe. Starting from first principles, we will develop the basics of metric differential geometry. Then the instructor will present an outline of Einstein's special and general theory of relativity, which describe the kinematics and dynamics of spacetime, and discuss aspects of black holes, gravitational lensing and optical geometry. Thanks to recent breakthroughs in astrophysics, such as the observation of a black hole shadow and gravitational waves, this is a subject of great current interest with many research opportunities. You will gain an overview of the field and learn mathematical techniques to help you develop your interests and implement Signature Work projects in this area. During the course of the week, we will shift from an intensive lecture-based to a more discussion-based format as participants carry out project work. This work will consist of guided reading of actual research papers in the field, and their presentation to the whole class on the final day. The course will also include a guest lecture by a distinguished researcher.
Prerequisites: MATH201 Multivariable Calculus and MATH202 Linear Algebra
Instructor: Marcus Werner
The Planetary Sensorium – Artistic Explorations of a Networked World
In this course, we introduce new emerging concepts and explorations within global digital arts discourse that may help you identify research questions and directions for your Signature Work project. We will explore the concept of networked systems through the philosophical and artistic discussions surrounding the new planetary realities that we are facing today, where the digital virtual world and the real material world are increasingly merging and intertwining, generating complex hybrid realities. We pose the question: How can art and artistic research provide a means for understanding and breaking down wicked problems? Throughout the mini term we will dissect cutting-edge interdisciplinary ideas that are being presented in the international art dialogue. We will explore how these strategies, precedents and research are allowing the arts to connect with other knowledge systems in tackling emerging complex, multifaceted issues. We will work in a studio environment that encourages experimentation with diverse modes of media, communication and expression that is and is not currently offered in existing courses within media arts along with supplemental lectures, guest presentations from artists, ample time to review and discuss your project ideas, brainstorming sessions and critique.
Instructors: Benjamin Bacon and Vivian Xu
What Could/Should Curating Do?
Are you curious about how an art exhibition is conceptualized, organized and presented? Are you interested in pursuing a career in curation or other related fields in contemporary art? This mini-term Signature Work course will provide you with a platform to intellectually and professionally engage with the art world. This course will give you a rare opportunity to directly work with “real-world” artistic and cultural practitioners, as well as ample time to explore your Signature Work design in the areas of art history, media and art, cultural studies and history, among others. For the culmination of the course, we will work on a collective exhibition or intervention on campus. The course will be in close collaboration with What Could/Should Curating Do? (WCSCD), an educational-curatorial platform initiated by the internationally acclaimed curator Biljana Ciric. The WCSCD program will be conducted in online format first with mentors from all over the world. The four-day mini-term course in March 2021 will be conducted by professor Zairong Xiang and one of the other international mentors of WCSCD.* We will use mornings for lectures and workshops, afternoons for individual meetings and group works, and evenings for fun – what is art if it is not fun?
Prerequisites: As a Signature Work mini-term course participants should have already taken at least two relevant courses in Arts and Humanities (for example Art History and Global Art History)
Instructor: Zairong Xiang (and invited mentors)
*The mini-term course forms part of the WCSCD, which lasts three months. Students in the mini-term course are NOT expected/required to follow the whole three-month program. This year, WCSCD provides up to five places for DKU students to enroll free of charge in the whole program with a selected group of emerging curators from India, Canada, the U.S., Central Asia, and Australia. Program participants will conduct interviews, engage in archival practices around their research into these practitioners, all while thinking about modes of archiving in and of itself and discussing different curatorial histories. Students interested in pursuing the whole WCSCD program should contact professor Zairong Xiang, associate director of arts, at firstname.lastname@example.org no later than Nov. 27, 2020.
Writing the Archive: An Exploration of Sources in Signature Work
What are our responsibilities toward the sources we use in research? What techniques can we use to write in— and also with, through, and about—the archive? This course invites students to explore the archive as a broad conceptual category in its many iterations: not only library papers but also family documents, digital repositories, newspaper and magazines collections, and so forth. We emphasize the ways in which research is shaped by institutions that aren’t neutral and by the researcher’s own set of biases and perspectives. During the course, we will look at different generic (and cross-genre) models of written responses to archival material: academic prose, personal narratives, poetry, collage, and more. The course time will consist of short readings, activities, and workshops, with students working independently and together to “write” their archives, generating material toward their Signature Work projects.
Prerequisites: It is highly recommended you contact the instructors beforehand to identify an “archive” of material you will explore in connection to your Signature Work project
Instructors: Stephanie Anderson and Caio Yurgel
Wrong but Useful Models
All models are wrong, but some are useful - the models are wrong because they heavily rely on assumptions which sometimes may not hold; the models can be useful because they help us to uncover causal relationship from observed data. This course will touch on the mainstream causal-inference models in social science, which can be used to carry out advanced data analysis in signature works. After taking this course, students will develop a better understanding of quantitative methods in social science, and will be able to generate research hypothesis that can be examined by causal inference models. To that end, students will be exposed to academic papers using quantitative analysis, and will discuss research proposals with the instructor as well as their peers. During this course, students will also learn how to find data and use statistical tools (e.g. STATA) to analyze them. This course may appeal to many students interested in economics, political science, sociology, public policy and data science.
Pre-requisite: STAT 101 (ECON 203 Econometrics is preferred but not required)
Instructor: Xiaochen Zhang
“You Can’t Cure Stupid”: The Origins of Medicinal Drugs
Medicinal chemistry is a branch of science that deals with the discovery and design of new therapeutical chemicals and their development into useful medicines. It is an area of study that is highly interdisciplinary, involving contributions from a number of fields represented at DKU. Most drug discovery programs involve a team of scientists including biologists, toxicologists, pharmacologists, theoretical chemists, and microbiologists, while the decision to pursue a drug development program involve economics and public policy considerations. The course will present students with the opportunity to engage with potential Signature Work ideas in medicinal chemistry within the DKU context. Using specific case studies involving some of the most important drugs, the development or discovery of pharmaceuticals will be described with respect to the conception process and techniques involved. Given the interdisciplinarity of this field, a wide range of areas from chemistry (inorganic, organic, analytical, physical) to biology, physics, material science and even fields within the Global Health major, will be obvious as potential topics that students could explore. Students in the social sciences may also have relevant academic interests, such that exposure to drug development will benefit their research interests.
Instructor: Floyd Beckford