The seminar organized by the Department of Knowledge Engineering on May 25, 2023, attracted the attention of students from the Faculty of Information Technology.
The seminar shared various methods of attacking machine learning models, as well as introduced techniques to ensure privacy during the training process. The attack methods included data poisoning attacks, adversarial attacks, and privacy attacks. These attack methods aim to manipulate machine learning models or exploit sensitive information in training data. The techniques to ensure privacy are built based on different phases of data usage, including data anonymization, differential privacy, homomorphic encryption, and federated learning. These techniques can be used individually or in combination to ensure privacy for training data and have been widely adopted in many modern technologies, with potential for further improvements in the future.
At the end of the seminar, students gained valuable knowledge in the field of Knowledge Engineering.
We sincerely thank Mr. Bui Huy Thong, the representative of the Department of Knowledge Engineering, for taking the time to attend the seminar and share insightful lessons.