Secure multiparty computation (MPC) is one of the most impressive achievements of modern cryptography, enabling distributed computations by mutually distrusting parties in a way that guarantees privacy of the inputs and correctness of the output. Broadcast can be viewed as a specific MPC task that allows a designated sender to reliably send a message to all other parties.
In recent years there has been an increasing gap between the success of machine learning algorithms and our ability to explain their success theoretically. Namely, many of the problems that are solved to a satisfactory degree of precision are computationally hard in the worst case. Fortunately, there are often reasonable assumptions which help us to get around these worst-case impediments and allow us to rigorously analyze heuristics that are used in practice.
Most human traits of interest are highly “polygenic” (or “complex”): the bulk of the heritable variation is due to a combination of numerous genetic variants of small marginal effects. A polygenic score is a predictor of a person’s complex trait value computed from his or her genotype. Polygenic scores sum over the genetic effects of the alleles carried by a person—as estimated in a genome-wide association study (GWAS) for the trait of interest.
Cloud computing opens a new chapter in information technology, by enabling global access to shared pools of resources such as services, data, servers, and computer networks. It drives new digital businesses across enterprises. In the last few years, an unprecedented amount of data center capacity has been built to support cloud computing services' growth.
We are living in the era of massive data. Some estimate that more than 90% of the worlds data was generated in recent years. This large amount of data comes with technological challenges such as: distributed storage systems, synchronization issues, density constraints, power constraints and more.
The talk will be partitioned into two, independent parts in which some solutions to those challenges will be discussed. Specifically, we will touch upon dynamical distributed storage systems, and Polya string models.
Satisfiability Modulo Theories (SMT) solvers are general-purpose engines that can decide whether a given set of constraints is satisfiable.
Their expressiveness and efficiency make them well-suited for many applications, and they have been, and continue to be successfully adopted by both industry and academia.
I will start this talk by briefly describing what is the SMT problem, and why it is a useful one. Then, I will present several works that either use SMT-solvers or improve them (or both) to increase the level of trust in three application domains.
I will survey two projects that study the theoretical properties of graph games and their applications. The first part of the talk focuses on the controller-design problem. Since in order to deploy a controller in practice, it is critical to guarantee both performance and correctness, the solution uses a combination of formal methods and machine-learning techniques. Specifically, we combine deep reinforcement learning with formal methods for graph games, towards a design of controllers that achieve high-quality performance and are provably correct.
My research focuses on dealing with core challenges of Computer Vision (CV) and applied Machine Learning (ML) to develop generalizable solutions for processing and analysis of complex data. I develop computational models to deal with high variability of image statistics, insufficient amount of labeled data, unbalanced data, and data normalization.
The design of correct and efficient multi-process systems, in which processes are executed in parallel, and influence each other, is a significant challenge for software engineers. In this talk, I will survey two of my research projects, that tackle this aim from different aspects. The first part concerns process-synchronization. I will present the signaling problem, proposed by Marcus Aguilera, Eli Gafni, and Leslie Lamport. Then, I will provide a solution to that problem by presenting an efficient signaling algorithm, as well as applications thereof.
Age-structured multiscale models have been developed to study viral kinetics. However, they are notoriously difficult to solve and when utilizing this type of models parameter estimation presents a challenge. Here, we investigate the numerical solutions of a multiscale model of hepatitis C virus (HCV) dynamics during antiviral treatment and compare them with analytical approximations.
