学术报告：Online Optimization for Network Resource Allocation and Comparison with Reinforcement Learning Techniques

报告时间：5月21日（星期二）下午14:00-15:00

报告地点：学院南路校区，主教213

报告人：赵以强，加拿大卡尔顿大学，教授

报告摘要：In this talk we introduce online algorithms for network resource allocation. The network considered is composed of many servers connected by communication links. The system operates in discrete time; at each time slot, the network administrator reserves resources at servers for future job requests with a cost. Then, after receptions of the resource demand, the jobs may be transferred between the servers to best accommodate the demands with an additional transport cost. Finally, if a job request cannot be satisfied, then there is a violation cost for the blocked job. We propose a randomized online algorithm based on the exponentially weighted method. We prove that our algorithm enjoys a sub-linear in time regret, which indicates that the algorithm is adapting and learning from its experiences, and becomes more efficient in its decision-making as more data become available. Moreover, we tested the performance of our algorithm on simulated data and compare it against a reinforcement learning method. The results show that our proposed method outperforms the reinforcement learning method.

报告人简介：赵教授曾获南京信息科技大学学士学位，加拿大萨斯喀彻温省大学数学与统计学系博士学位，于2003年起任职卡尔顿大学教授，现担任数学和统计学学院的副院长。