Yanfang (Fanny) Ye, assistant professor of computer science and electrical engineering at West Virginia University, has been awarded a grant from the National Science Foundation to enhance security for machine learning mechanisms. The award comes with more than $237,000 in funding over a three-year period.
As artificial intelligence and machine learning techniques have been widely used for cyber defense purposes such as malware and fraud detection, the incentive for defeating them via intelligent evasion attacks increases. Evasion attacks cause the model to misclassify a sample in order to evade detection.
“There are currently no effective countermeasures against these sophisticated attacks,” said Ye. “The objective of this project will be to investigate effective ways to make machine learning techniques such as classification and clustering mechanisms robust against these types of attacks.”
Ye will investigate a more powerful class of attacks known as gray-box attacks, which allow the attacker to perform all the activities that a defender would normally perform. She will then build a theoretical model for characterizing the vulnerability and resilience of machine learning mechanisms with respect to intelligent evasion attacks under the gray-box model, thus enhancing their security to withstand these attacks with quantifiable resilience gains.
“This project will also involve doctoral students who will directly contribute to the next-generation workforce,” Ye said, “and will address diversity by involving female students and students from underrepresented groups.”
Ye has extensive research and development experience in the cybersecurity industry. Before joining WVU, she was the principal scientist in Comodo Security Solutions, Inc., a provider of computer software and SSL digital certificates in the U.S., and deputy director at Kingsoft Internet Security Corporation, the second biggest Internet security company in China. Ye proposed and developed cloud-based solutions for mining big data in the area of cybersecurity, especially for malware detection and phishing fraud detection. Her developed algorithms and systems have been incorporated into popular commercial products, including Comodo Internet Security and Kingsoft Antivirus that protect millions of users worldwide.
She received the prestigious ACM SIGKDD 2017’s Best Paper and Best Student Paper awards (Applied Data Science Track), the IEEE EISIC 2017 Best Paper Award and the 2017 New Researcher of the Year Award from the Statler College. Ye has brought in more than $1.35 million in research funding from the NSF in the past two years.