Zuckerman AI Fund at Technion

AI Impact: The Zuckerman Fund for Interdisciplinary Research in Machine Learning and Artificial Intelligence at the Technion

Zuckerman AI Fund at Technion

The Technion’s Center for Machine Learning and Intelligent Systems (MLIS) was established to address relevant challenges in artificial intelligence and machine learning science, intelligent systems, and technology.

As part of this AI initiative, The Zuckerman Fund for Interdisciplinary Research in Machine Learning and Artificial Intelligence was created to tap into the Technion’s record of advancing ground-breaking interdisciplinary research and to place Israel at the center of the global AI map. With 100 faculty members engaged in research across the AI spectrum, the Technion has become an academic leader in advancing AI, both in Israel and around the world.

The Technion Center for MLIS is co-directed by Profs. Shie Mannor and Assaf Schuster of the Technion’s Henry and Marilyn Taub Faculty of Computer Science. More than 100 affiliated staff members reflect the “core” of AI-related faculties (Computer Science, Electrical Engineering, and Industrial Engineering) as well as “user” faculties (Biomedical Engineering, Mechanical Engineering, and Aerospace Engineering).

MLIS acts as an umbrella and focal point for all Technion AI activities, research, and collaboration with industrial partners.

MLIS projects are led by the following researchers:
Asst. Prof. Ido Kaminer (Faculty of Electrical and Computer Engineering) designed the Ramanujan Machine, a novel “conjecture generator” that creates mathematical conjectures, which are considered the starting point for developing mathematical theorems.

Dr. Or Aleksandrowicz from the Architectural Research Lab (Faculty of Architecture and Town Planning) led a project based on the collection and analysis of Big Data, using multidisciplinary research related to architecture and urban environments, to support the development of advanced building technologies.

Dr. Yaniv Romano (Faculty of Electrical and Computer Engineering) is leading a machine-learning project focused on design learning and statistical methodologies to effectively identify explanatory features (e.g., genetic variations) linked to a phenomenon under study (e.g., disease risk), while rigorously controlling the number of false positives among the reported features.

Asst. Prof. Daniel Soudry (Faculty of Electrical and Computer Engineering) is carrying out research addressing the core challenges of (1) understanding deep learning and (2) making it more efficient in terms of computational resources.

Prof. Roi Reichart (Faculty of Industrial Engineering and Management) has conducted research into Natural Language Processing (NLP) focused on language learning and design models that integrate domain and world knowledge with data-driven methods.

Assoc. Prof. Vadim Indelman’s Autonomous Navigation and Perception Lab (ANPL, Faculty of Aerospace Engineering) investigates problems related to single and multi-robot collaborative autonomous navigation and perception, with a particular focus on accurate and reliable operation in uncertain environments.

Prof. Assaf Schuster is leading the Asynchronous Distributed Training of Deep Neural Networks Project, which has developed asynchronous versions of data-parallel training and showed them to be faster than their synchronous counterparts, contributing to efficient cloud computing.

Collaboration with the Grand Technion Energy Program (GTEP)

In 2021, MLIS, in collaboration with the Grand Technion Energy Program (GTEP), initiated a joint call for proposals, seeking to promote intra-Technion collaborative research in the fields of AI/ML and energy.

Winning Research Proposals:

Deep Learning for Solar-To-Chemical Energy Conversion
Prof. Lilac Amirav, Dr. Kira Radinsky
This project utilizes machine-learning, based on a novel integrated poly-cycle architecture, for guided search of alternative and improved organic molecules to be oxidized simultaneously with H2 production, using a unique and highly efficient nanorod photocatalytic system, for direct solar-to-chemical energy conversion.

Development of an Urban-Scale Building Electricity Consumption Model Based on Building Characteristics, Climatic Conditions and Occupant Behavioral Profiles
Prof. Avigdor Gal, Dr. Or Aleksandrowicz
This research project analyzes actual large-scale electricity consumption data gathered from Israel Electric Corporation (IEC) meters. The data will be used to develop a model for predicting energy consumption in buildings, while considering the physical characteristics of the existing buildings’ stock, climatic and meteorological data, and electricity usage profiles.

Improving the Trustworthiness of Machine Learning Algorithms for Power System Applications using Explainable Artificial Intelligence Techniques
Prof. Yoash Levron, Prof. Shie Mannor, Dr. Kfir Yehuda Levi
The main goal of this research is to develop methods to explain the results of ML algorithms used in power system applications.
 

 

Zuckerman Institute Support of the Martin and Grace Druan Rosman High-Performance Center (HPC)

Thanks to the support of the Zuckerman Institute, the Technion completed its inital upgrade of the HPC in 2021. The final phase of the upgrade, scheduled to be completed in the first quarter of this year, will dramatically improve the Center’s energy capacity, making it the most powerful computing cluster in Israeli academia and a center for excellence in calculation-based research.

HPC Building

HPC Building