New at the Faculty: Deep Learning Lab

In the upcoming academic year, the Faculty of Engineering will feature a new lab in the field of deep learning and its applications. The lab, headed by Dr. Yoli Shavit, will focus on developing innovative methods for handling complex and expansive data, offering solutions to challenges of generalization and uncertainty.
The Age of Data and the Need for Deep Learning
In today's age, data has become the most valuable resource. From incredible quantities of time series data from various industries, through complex biological models, to smart localization systems—the ability to extract insights and make automatic decisions from this data is critical. However, their enormous complexity and unprecedented scale make these tasks particularly challenging for humans. This is where deep learning comes into play.
Deep learning has already proven its worth in a wide range of applications, from computer vision and natural language processing to precision medicine. However, many of these technologies still encounter difficulties in solving problems for complex applications. For example, analyzing time series with multiple dependencies, building biological models with countless variables and relationships, and performing precise localization in dynamic environments. The limited ability of deep learning models to quantify the uncertainty of their predictions also constitutes a significant barrier for critical applications, where any error could have grave consequences
Laboratory Goals: Innovative Solutions for Complex Challenges
Dr. Shavit's research group will focus on developing advanced algorithms and architectures in deep learning that will address these challenges. In particular, emphasis will be placed on methods that integrate multiple data modalities (multi-modality) and uncertainty quantification. The key goal is to substantially improve model performance and their generalization capabilities, and to enable more reliable decision-making based on their outputs.
Last Updated Date : 31/07/2025