Tuvia Newman

Title: Tuvia Newman: Cybersecurity and Machine Learning Inventor

Introduction

Tuvia Newman is an inventor based in Kiryat Gat, Israel. His patent work focuses on applying machine learning and artificial intelligence to cybersecurity, including methods for assessing cyber incidents and prioritizing security alerts. He is associated with Palo Alto Networks, Inc. Newman has 3 patents.

Latest Patents

One of Newman’s latest patents, “Training a Model Based on Soft Labeling,” describes a cybersecurity method for training a machine learning model. The method uses a corpus of cyber incidents, their features, and binary labels indicating whether each incident is benign or malicious. Predefined rules map those binary labels to soft labels that represent levels of suspiciousness. The incidents and soft labels are then used to train a model that can predict risk scores for incidents outside the training corpus.

Another patent, “Adaptive Incident Prioritization Based on User Feedback,” describes a computer-implemented approach to ranking security incidents. Pre-trained AI models analyze features extracted from incidents and produce a prioritization for a particular computer system. A user reviews the prioritization and provides feedback about its quality. The system uses that feedback to adjust how it extracts features for subsequent prioritization.

Career Highlights

Newman’s disclosed inventions address two complementary cybersecurity challenges: training models to assess incident risk and adapting incident prioritization to user feedback. Together, these approaches illustrate how machine learning can support security teams in evaluating and organizing incidents.

Collaborations

Newman is associated with Palo Alto Networks, Inc. His listed coworkers include Yinnon Meshi and Gal Itzhak.

Conclusion

Tuvia Newman’s patent portfolio highlights his work on machine-learning methods for cybersecurity, with an emphasis on risk scoring and user-informed incident prioritization.