Semi-Supervised Time Series Anomaly Detection Based on Statistics and Deep Learning

  • Event Date: 2025-05-21
  • Quantum information and communication
  • Speaker: Prof. Jehn-Ruey Jiang (Department of Computer Science and Information Engineering, NCU)  /  Host: Prof. Li-Yi Hsu (CYCU)
    Place: R211, Science Building, Chung Yuan Christian University

Time:14:00, Wednesday, May 21, 2025
Speaker:Prof. Jehn-Ruey Jiang
              Department of Computer Science and Information Engineering, NCU
Title: Semi-Supervised Time Series Anomaly Detection Based on Statistics and Deep Learning
Place : R211, Science Building, Chung Yuan Christian University

Abstract:
In this talk, I will first introduce the concept of the graphics processing unit (GPU). Next, I will introduce a GPU annealer (GPUA), a quantum-inspired annealer based on the GPU. This annealer utilizes diverse adaptive bulk search (DABS) to find  the optimal solution to combinatorial optimization problems (COPs). DABS is essentially a genetic algorithm offering eight genetic operations and five local search algorithms to enhance annealer performance. At the end of the talk, I will present benchmark comparisions among a quantum annealer (QA), a digital annealer (DA), and a GPUA. 

1/0