Multitask Active Learning for Graph Anomaly Detection

1 code implementation • 24 Jan 2024 • Wenjing Chang, Kay Liu, Kaize Ding, Philip S. Yu, Jianjun Yu

Firstly, by coupling node classification tasks, MITIGATE obtains the capability to detect out-of-distribution nodes without known anomalies.

Active Learning Graph Anomaly Detection +2

Real-time Dual-channel 2 * 2 MIMO Fiber-THz-Fiber Seamless Integration System at 385 GHz and 435 GHz

no code implementations • 24 Jun 2022 • Jiao Zhang, Min Zhu, Bingchang Hua, Mingzheng Lei, Yuancheng Cai, Liang Tian, Yucong Zou, Like Ma, Yongming Huang, Jianjun Yu, Xiaohu You

We demonstrate the first practical real-time dual-channel fiber-THz-fiber 2 * 2 MIMO seamless integration system with a record net data rate of 2 * 103. 125 Gb/s at 385 GHz and 435 GHz over two spans of 20 km SSMF and 3 m wireless link.