RESEARCH

Artificial Intelligence has emerged as a powerful approach that has already led to many breakthroughs in multiple scientific and technological fields in the past decade.

Deep Learning, inspired by the layered and hierarchical deep architecture of the human brain, has emerged as an efficient means to design photonic structures.

Our current proposal stems from the complementary expertise of the two groups and from the realization that Deep learning strength can be further leveraged to address inverse design challenges in physics and, in particular, in electromagnetism’s sub-wavelength inverse design spanning the large spectrum of electromagnetic and photonics applications.

More specifically, using recently introduced DL methods and our recent achievements, we plan to address challenging inverse design tasks in electromagnetism such as sensing with nano-photonics, novel 5G and Wifi antennas as well as integrated photonics for quantum information and LIDARs.

Social Networking Sites (SNSs) have become the new town hall. We get much of our news through SNSs, and this is where we hold discussions and debates. Therefore, SNSs have large influence on opinions and decisions.

At the heart of these SNSs there are recommendation systems that tailor the feed to our preferences. These systems use machine learning algorithms to make the feed more to our liking.  

Therefore, these machine learning algorithms shape the way we perceive the world. The goal of this study is to answer the fundamental question: Do SNSs actively change people opinions or, alternatively, do the SNSs just reflect users’ opinions to them? 

Instrumental variable (IV) estimation is a widely used method that supports high stakes government policies and business decisions, when controlled experiments are not feasible or when a treatment is not successfully delivered to every unit in a randomized experiment.

In this research project, we will develop statistical methods for assessing the validity of such designs. IV designs rely on exclusion restriction assumptions that are not directly testable and that are therefore challenging to assess empirically. In this context, this project aims to formalize the logic of falsification tests, a set of tests that leverage contextual knowledge about the absence of causal links (for example, from future to past outcomes) to test the validity of candidate instruments.

We will establish that IV falsification tests can be mapped to a class of prediction problems that can leverage current machine learning methods. Based on this conceptualization, we will develop general methods that would both improve falsification test efficiency and help guide the construction of such tests.

These methods will be particularly applicable to research using large datasets with many candidate variables that can be used for falsification, an increasingly common situation for which no formal methods currently exist. Developing methods for evaluating and improving the validity of these common research designs entails clear societal benefits.