首页|China Agricultural University Reports Findings in Support Vector Machines (MTKSV CR: A novel multi-task multi-class support vector machine with safe acceleration rule)
China Agricultural University Reports Findings in Support Vector Machines (MTKSV CR: A novel multi-task multi-class support vector machine with safe acceleration rule)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – New research on Support Vector Machine s is the subject of a report. According to news originating from Beijing, People ’s Republic of China, by NewsRx correspondents, research stated, “Regularized mu lti-task learning (RMTL) has shown good performance in tackling multi-task binar y problems. Although RMTL can be used to handle multi-class problems based on ‘o ne-versus-one’ and ‘one-versus-rest’ techniques, the information of the samples is not fully utilized and the class imbalance problem occurs.” Our news journalists obtained a quote from the research from China Agricultural University, “Motivated by the regularization technique in RMTL, we propose an or iginal multi-task multi-class model termed MTKSVCR based on ‘one-versus-one-vers us-rest’ strategy to achieve better testing accuracy. Due to the utilization of the idea of RMTL, the related information included in multiple tasks is mined by setting different penalty parameters before task-common and task-specific regul arization terms. However,the proposed MTKSVCR is time-consuming since it employ s all samples in each optimization problem. Therefore, a multi-parameter safe ac celeration rule termed SA is further presented to reduce the time consumption. I t identifies and deletes most of the superfluous samples corresponding to 0 elem ents in the dual optimal solution before solving. Then, only a reduced dual prob lem is to be solved and the computational efficiency is improved accordingly. Th e biggest advantage of the proposed SA lies in safety. Namely, it derives an ide ntical optimal solution to the primal problem without SA. In addition, our metho d remains effective when multiple parameters change simultaneously.”
NETL
NSTL
