Prof. Stern is an associate professor at Tel Aviv University. She obtained her bachelor’s degree in Biology, Psychology, and an additional degree in Mathematics from Tel- Aviv University, where she also went on to complete her M.Sc. and Ph.D. in probabilistic modeling of evolutionary processes. During her postdoctoral research at the Weizmann institute she pioneered some of the first genomics studies on CRISPR and phage co-evolution. She then continued to a second post-doc jointly at UCSF and UC Berkeley, probing the evolution of polioviruses. In her work she combines mathematical modeling, analysis of massive sequencing data, and experiments that reveal how RNA viruses evolve. Since 2014 she is at the Shmunis School of Biomedical and Cancer Research, where she has led several breakthrough studies on COVID-19 and other pathogenic viruses Prof. Stern has won numerous prizes and awards, including a UNESCO-L’Oreal prize for women in science, the Wolfe fellowship for distinguished students, and the Rothschild postdoctoral fellowship. Dr. Stern had been listed first as one of the top ten brightest graduate students in Israel, and recently as one of the 70 brilliant leading women in Israel.
 

Drug development process is a lengthy and expensive process that can take more than 10 years and cost billions of dollars. Despite the major efforts, the process will fail in more than 90% of the cases due to safety concerns.
In this work we present a different approach for the task of assessing drug safety. An alternative to the animal models widely used today. We tackle the task using a BIO-AI approach. The main ingredient is a 'patient-on-a-chip' platform. The platform acts both as a human body simulator and as a data generator. The platform is capable of performing millions of interactions between the patient-on-a-chip and potential drugs, the data generated using the platform is used to train multimodal deep learning models to predict drug safety. Trained models are used by big pharma companies to predict safety of new drug candidates, and to provide personal safety recommendations for using existing drugs on specific patients.

Shahar Harel is the the Head of AI at Quris, where he tackles some of the world's greatest challenges in the drug development space. Previously to Quris, he led data science teams both in NASDAQ-listed companies and in smaller, highly innovative startups. Shahar holds an M.Sc. in Computer Science from the Technion – Israel Institute of Technology. During his research with Dr. Kira Radinsky, Shahar developed generative deep learning methods for generation of novel chemical compounds with unique characteristics, with the ultimate goal of accelerating drug discovery and development. His research works are presented at high profile data science conferences and published in top-tier pharmaceutical science journals.

Deep learning has revolutionized the field of machine learning with ground-breaking applications in various domains, including image processing, text analysis, and speech recognition. Deep learning has also become ubiquitous in bioinformatics and genomics. In this short talk, I will describe one of our sweeping successes in applying deep neural networks to high-throughput genomics datasets, and how we interpret the trained neural networks to discover new biology.

Yaron Orenstein is an Associate Professor of Computer Science and Life Sciences and the head of the Computational Biology lab at Bar-Ilan University. Yaron completed his BSc summa cum laude in Electrical Engineering and Computer Science at Tel-Aviv University, where he continued on a direct MSc track under the supervision of Prof. Dana Ron. He then completed his PhD in Computer Science at Tel-Aviv University supervised by Prof. Ron Shamir, where he received numerous awards and fellowships, such as the Deutch prize and the Dan David fellowship. He completed his post-doctoral training at the Massachusetts Institute of Technology with Prof. Bonnie Berger, and spent a semester as a Research Fellow at the Simons Institute for the Theory of Computing at UC Berkeley. In the last five years, Yaron has been the head of a fruitful and productive lab with numerous publications, grants, and graduate students. He authored more than 40 journal manuscripts and conference proceedings papers, received grants from the ISF, BSF, NIH, ICA, and IIA, and mentored more than 15 graduate students. His main research interests include sequence design problems and the application of deep neural networks in genomics.

The following two fields have barely interacted before: reinforcement learning and molecular evolution. Here we develop a reinforcement-learning algorithm to tackle the challenge of reconstructing phylogenetic trees, which are used to describe the evolutionary relationships among a set of genome sequences. When only a few dozen of sequences are analyzed, billions of alternative phylogenetic trees could potentially describe the evolutionary patterns, thus rendering the search for the tree that best describes the data algorithmically challenging. Thus, all current algorithms for phylogenetic tree reconstruction use various heuristics to make tree inference feasible. Although recent studies have shown the potential of harnessing AI-based methods to reconstruct phylogenetic trees, the question regarding the overall tree-search strategy remains open. In this study, we provide a novel paradigm for predicting the maximum-likelihood tree by using reinforcement-learning techniques to learn an optimal strategy for tree search. We demonstrate the use of the developed deep-Q-learning agent on a set of unseen empirical data, namely, on unseen RL environments defined by nucleotide MSAs of up to 20 sequences. Our results show that the likelihood scores of the inferred phylogenies is as accurate as those obtained from widely-used software. It thus establishes a proof-of-concept that it is beneficial to optimize a sequence of moves in the search-space, rather than only optimize the progress made in every single move. This suggests that an RL-based method provides a promising direction for phylogenetic reconstruction.

I research the application of advanced learning techniques on  problems from biological domains. Once the recent state of the art reinforcement learning algorithms were developed, my interest in the field grew and became my main research interest. As part of Prof. Tal Pupkos group and together with Dana Azouri from Itay Mayrose's group and Oz Granot from Ishay Mansour's group I've been working on modeling various biological problems as games and training agents to solve them.
I hold a bachelor's degree in Bioinformatics from Tel Aviv University and am about to finish my Master's in the same field.
 

We propose a new method to probe the learning mechanism of Deep Neural Networks (DNN) by perturbing the system using Noise Injection Nodes (NINs).
These nodes inject uncorrelated noise via additional optimizable weights to existing feed-forward network architectures, without changing the optimization algorithm.
We find that the system displays distinct phases during training, dictated by the scale of injected noise.
We first derive expressions for the dynamics of the network and utilize a simple linear model as a test case. 
We find that in some cases, the evolution of the noise nodes is similar to that of the unperturbed loss, thus indicating the possibility of using NINs to learn more about the full system in the future. 

Noam Levi is a Ph.D. student in Theoretical Particle Physics at Tel Aviv University. His academic research focuses on the theoretical understanding of the fundamental constituents of the universe as well as the theoretical and algorithmic foundations of deep learning. Noam earned a B.Sc. in electrical engineering and a B.Sc. in physics at Tel Aviv University, as well as an M.Sc. (magna cum laude) in theoretical particle physics, and is currently pursuing his Ph.D. under the supervision of Prof. Tomer Volansky. During his Ph.D., he spent two years conducting research at the Institute for Advanced Study in Princeton. Noam is an inaugural Milner foundation fellow, awarded the “70 for 70” Ph.D. scholarship. 

Dimensionality reduction has become an indispensable tool in the modern algorithmic toolbox, often being used as a preprocessing step in data analysis and machine learning pipelines. Rigorously analyzing quality of such methods is therefore of great importance for applied community as well as of fundamental theoretical interest.
In this talk I will focus on metric dimensionality reduction, the task of embedding high-dimensional datasets into a small number of dimensions and preserving the metric structure of the input.
A rich body of research in Math and CS has been devoted to developing reliable and efficient embedding methods, with most techniques designed to optimize the worst-case quality guarantees. That is, the objective is to approximate all the pairwise distances of the dataset points. Yet much of this research line has little relevance to practical settings. Indeed, in nearly all practical applications of metric dimensionality reduction the fundamental criterion for measuring quality is its average performance over all pairs. Preserving distances among most of the pairs in the data will usually suffice for a successful application.
Various average-case quality measures have been suggested and studied from a practical point of view, and many heuristics have been developed to optimize these notions. However, almost nothing is known in terms of rigorous analysis and absolute bounds on the quality of these methods.
In the talk, I will discuss some of the most common practically oriented average-case quality measures. I will then focus on dimensionality reduction for Euclidean datasets and present our theorem stating that the good old random projection method is, surprisingly, the best possible for these measures. For the general datasets, I will present the first approximation algorithm with provable guarantees for the average quality measurement criteria and will conclude with several experimental results showing a clear advantage of the random projection-based method over two popular heuristics, PCA and ISOMAP.

Dr. Ora Fandina is a Research Fellow at the ML & Algo group in the CS Department of Aarhus University in Denmark. Her research interests include unsupervised learning, clustering, optimization, and approximation theory. A primary focus of Ora's research is developing accurate, efficient, and practical methods for dimensionality reduction, as well as exploring their computational limits. She received her Ph.D. from the CS department at Hebrew University in 2020, and her MSc and BSc from the CS department at Ben-Gurion Univerisy in 2013 and 2010 respectively. In addition to research, Ora is dedicated to teaching, mentoring, and  encouraging young women to study math and science.

The lower the distortion of an estimator, the more the distribution of its outputs generally deviates from the distribution of the signals it attempts to estimate. This phenomenon, known as the perception-distortion tradeoff, has captured significant attention in signal restoration, where it implies that fidelity to ground truth signals comes at the expense of perceptual quality (e.g. deviation from statistics of natural images). However, despite the increasing popularity of performing comparisons on the perception-distortion plane, there remains an important open question: what is the minimal distortion that can be achieved under a given perception constraint? In our work, we derive a closed form expression for this distortion-perception (DP) function for the mean squared-error (MSE) distortion and the Wasserstein-2 perception index. We prove that the DP function is always quadratic, regardless of the underlying distribution. This stems from the fact that estimators on the DP curve form a geodesic in Wasserstein space. In the Gaussian setting, we further provide a closed form expression for such estimators. For general distributions, we show how these estimators can be constructed from the estimators at the two extremes of the tradeoff: The global MSE minimizer, and a minimizer of the MSE under a perfect perceptual quality constraint. The latter can be obtained as a stochastic transformation of the former. Our theoretical results shed light on several topics that are subject to much practical activity, such as diverse image restoration and traversing between fidelity and visual quality during test-time, without training different models.

Dror Freirich received his B.Sc. in Computer Engineering from the Technion and M.Sc. degree in Electrical Engineering from Tel Aviv University, under the supervision of Prof. Emilia Fridman. He is currently finishing his studies towards a PhD degree at the Technion under the supervision of Prof. Ron Meir. His research interests include Control Systems, Reinforcement Learning and Machine Learning. His current work focuses on theoretical bounds in Estimation Theory, with applications to Signal Processing, Computer Vision, Control and ML.

The field of reinforcement learning has demonstrated remarkable results in recent years. One challenge in bringing these results into the real world is the high risk-sensitivity in many applications. In the lecture, we will discuss the algorithmic difficulties caused by risk-sensitivity in reinforcement learning, namely, sample inefficiency and blindness to successful strategies. We will present the novel CeSoR algorithm for risk averse reinforcement learning. In a driving problem, for example, CeSoR not only prevents all accidents, but also learns intuitive risk-averse driving policies.

The lecture is based on the paper "Efficient Risk-Averse Reinforcement Learning" (NeurIPS 2022).
 

Ido Greenberg is a machine learning researcher who focuses on fundamental understanding of AI algorithms and their applications. In addition to his experience as a machine learning consultant, he developed reinforcement learning chatbots at Google, robust meta learning systems at Nvidia, and trading algorithms at Istra Research. Ido completed his BSc in mathematics and physics in Talpiot Program at the Hebrew University. In his MSc in applied mathematics at Tel Aviv University, he studied algorithms for molecules reconstruction. He is currently pursuing his PhD in Electrical Engineering at the Technion.