Our School Team Won the Bronze in "Digital Intelligence Service Innovation Challenge Service Scheduling Competition" of Ali Tianchi


On December 5, 2020, the final of "Alibaba Cloud Digital Intelligent Service Innovation Challenge-Service Scheduling Competition" jointly organized by Alibaba and Zhejiang University was successfully held in Hangzhou. Under the guidance of Zhu Zongwei, the teacher of the Intelligent Cloud Computing Center of Suzhou Research Institute,University of Science and Technology of China, our representative team, which composed of second year master students Ding Juntao, Liu Weihong, Zhai Wenjie and Tang Xin, won the bronze medal in the competition.

Photo 1 Awarding Scene of the Competition

This service scheduling competition has won extensive attention from both the industry and academic field, and quite a few professors and industry elites have been invited to serve as judges, including Jin Li, vice president of Alibaba group; Professor Genke Yang, executive director of Ningbo Institute of artificial intelligence, Shanghai Jiaotong University; Professor Zhijiang Shao, Dean of School of control science and engineering, Zhejiang University; Chenghuang Shen director of Alibaba cloud intelligent global technology service department; Mengchang Wang, senior algorithm expert of machine intelligence technology of Dharma Institute;Yin Zhang,senior of Artificial Intelligence Department of newbird network Algorithm experts, etc

Photo 2 Our Team Representative Team


The contest problem comes from actual business scenario where AliYun Cloud provides full-cycle technical services to many companies, government agencies, and developers around the world. Among them, service scheduling optimization is one of the cores of improving customer experience. However, in face of immense service volume, largely differentiated regions and languages, diverse customer issues, complex service scenarios, and technical personnel with their own strengths, how to ensure the rapid distribution of massive service while being able t.o efficiently match service scenarios with the capabilities of technical experts is a difficult point in the service scheduling optimization algorithm. This service scheduling competition focuses on the combination of production, research and learning, and provides valuable experience for the development of "digital intelligence service" industry in China.

The competition has lasted three months and have attracted 1381 teams from more than 100 universities and enterprises, including Tsinghua University, Peking University, University of Science and Technology of China, Zhejiang University, Shanghai Jiaotong University, Nanjing University, Xi'an Jiaotong University, Huazhong University of Science and Technology, National University of Defense Technology, etc. Our team proposes a task scheduling method combining particle swarm optimization and shallow neural network. We constructed a new operator to replace the complex nonlinear relationship of deep neural networks and used particle swarm optimization to solve the difficult problem of BP back propagation in NP problem. Our team proceeded to the final since we won the sixth place in the preliminary round and topped at the third place in the second round. In the final, our team has been awarded the bronze medal with the total score ranking as the fourth.

Photo 3 Gathering of the Representatives of the Organizer and Contestants