Social scientists are increasingly turning to computer-assisted text analysis as a way of understanding the digital footprints left by communities and individuals. Much of the technology that powers these approaches is borrowed from the fields of computer science and statistics; yet, social scientists have substantially different goals. We focus on the development of methods that support three core tasks: discovery, measurement and causal inference with text. We introduce the Structural Topic Model (STM), a bayesian generative model of text which…

Abstract: Local geometric analysis is a method to define a coordinate system in a small neighborhood in the space of distributions over a given alphabet. It is a powerful technique since the notions of distance, projection, and inner product defined this way are useful in the optimization problems involving distributions, such as regressions. It has been used in many places in the literature such as correlation analysis, correspondence analysis. In this talk, we will go through some of the basic…

Abstract: As authoritarian governments around the world develop sophisticated technologies for controlling information, many observers have predicted that these controls would be ineffective because they are easily thwarted and evaded by savvy Internet users. In Censored, Margaret Roberts demonstrates that even censorship that is easy to circumvent can still be enormously effective. Taking advantage of digital data harvested from the Chinese Internet and leaks from China's Propaganda Department, this book sheds light on how and when censorship influences the Chinese…

  Abstract: Targeted socio-economic policies require an accurate understanding of a country's demographic makeup. To that end, the United States spends more than 1 billion dollars a year gathering census data such as race, gender, education, occupation and unemployment rates. Compared to the traditional method of collecting surveys across many years which is costly and labor intensive, data-driven, machine learning driven approaches are cheaper and faster--with the potential ability to detect trends in close to real time. In this work,…

Abstract: Methods for managing dynamic systems typically invoke one of two perspectives. In the worst-case perspective, the system is assumed to behave in the worst possible way; this perspective is used to provide formal safety guarantees. In the risk-neutral perspective, the system is assumed to behave as expected; this perspective is invoked in reinforcement learning and stochastic optimal control. While the worst-case perspective is useful for safety analysis, it can lead to unnecessarily conservative decisions, especially in settings where uncertainties…

  Abstract: Collaborative autonomous robots promise to revolutionize transportation, disaster response, and environmental monitoring. Already, micro-aerial vehicles have become a multi-billion-dollar industry; and in this new decade, teams of semi-autonomous ships, cars, and underwater exploration vehicles are being launched. A future of ubiquitous autonomy is becoming a reality, where robots can autonomously split into teams, and navigate and learn the world. However, this future is threatened by attacks and failures that can compromise the robots’ teams, control, and learning capabilities,…

  Abstract: Despite high-profile advances in various decision-making and classification tasks, Deep Neural Networks (DNNs) have found limited application in safety-critical domains such as self-driving cars and automated healthcare. In particular, DNNs can be vulnerable to adversarial attacks and input uncertainties. This issue becomes even more complicated when DNNs are used in closed-loop systems, where a small perturbation (caused by noisy measurements, uncertain initial conditions, disturbances, etc.) can substantially impact the system being controlled. Therefore, it is of utmost importance…

Abstract: The expansion of large-scale cyber-physical systems such as electrical grids, transportation networks, IoT, and other societal networks has created enormous challenges for controlling them and, at the same time, tremendous opportunities for utilizing the massive amounts of data generated by them. At the core of these data-driven control problems are distributed and stochastic optimization algorithms, such as the widely used stochastic gradient descent (SGD). While easy to use, these algorithms are not so easy to analyze, and in many…

  Abstract: Climate change demands faster material innovations in multiple domains to reduce the carbon emissions of various industrial processes, but it currently takes 10-20 years to develop a single material with conventional human-driven approaches due to the large amounts of trail and errors needed. Machine learning approaches enable the direct learning of structure-property relations of materials from data, which have the potential to accelerate the search of materials and understand new scientific knowledge. However, one central challenge in developing…

  Abstract: Solid-liquid interfaces have notoriously haphazard atomic environments. While essentially amorphous, the liquid has short-range order and heterogeneous dynamics. The crystal, albeit ordered, contains a plethora of defects ranging from adatoms to dislocation-created spiral steps. All these elements are of paramount importance in the crystal growth process, which makes the crystallization kinetics challenging to describe concisely in a single framework. In this seminar I will introduce a novel data-driven approach to systematically detect, encode, and classify all atomic-scale mechanisms…