One of the exciting recent developments in Machine Learning and AI is the success of Deep Learning algorithms to utilize big data towards solving some of the most emerging and important problems we face today.

This success, though, comes with a price. Often, machine learned models exploit information generated and owned by individuals that contains sensitive, private, copyrighted material.

It is then an emergent question how to harness these learning machines in a fair setting that enables the exploitation of the data, while not breaching the individual's rights.

In this proposal we focus on the question of generating synthetic data. We consider a setting where a learner has access to publicly available data, and may exploit it for fair-use to generate similar data, but the learner is not allowed to copy or to breach intellectual property (IP) rights.

The first challenge we face is to correctly model the question of IP-rights and fair use. In particular, the project will aim to propose different notions and criteria for fair-use as well as study their statistical and computational complexity.

The global marine environment is undergoing dramatic and rapid human induced changes. Alien and invasive species (AIS) are a major source of biodiversity loss and have caused major economic losses and threats to human health. While the magnitude of AIS in the Mediterranean is unprecedented, we lack simple and effective approaches to track AIS and understand resulting changes in ecological community structure and ecosystem dynamics. 

Here, we propose to use the wealth of information available in digital citizen science and social media repositories with a combination of automated image recognition and ecological modelling, to gain comprehensive insights on spatio-temporal dynamics of Mediterranean fish populations. By calibrating the tools in the Red Sea, the source of introduction to the Mediterranean, we will be able to detect new introductions at the initial stage, where densities are low and identification is difficult, providing a critical early warning for AIS.

Digital sources will be combined with a set of cameras along the Israeli coast to produce an automated tool that will continuously track changes, map results, and update range dynamic predictions. This will enable elucidating several key aspects of fish dynamics, such as undetected AIS introductions, incipient range expansion of AIS, migration routes, range size declines of Indigenous species, and changes in phenology associated with climate change.

This near real-time monitoring of fish dynamics will provide critical new information for scientists, policy makers, and the general public needed towards appropriate preparation for this emerging threat.

AI has the potential to revolutionize scientific research. Indeed, collaborations between data-scientists and researchers across the TAU campus have already resulted in unique AI-driven discoveries.

The aim of this project was to devise a comprehensive methodology for using deep-learning within all aspects of scientific and engineering research (e.g., simulations, inverse models, and interpretability).

The resulting methods will be applied to research at TAU.

Understanding the human brain is one of the greatest challenges in modern science. A key empirical tool in the arsenal is neuroimaging and specifically MRI.

The data generated by imaging studies are complex and their analysis requires the application of advanced DS methods, as well as the development of new options. In this project, we propose the creation and utilization of a neuroimaging database at TAU, which will be unique in terms of size (over 10000 subjects) and the multimodality of the collected data.

We will also form a collaboration of core DS researchers and neuroscience collaborators with the objective of designing new methodologies to drive scientific discovery.