This paper considers two parallel machine scheduling problems, where the objectives of both problems are to minimize the makespan, and the jobs arrive over time, on two uniform machines with speeds 1 and s （s 〉 1）,...This paper considers two parallel machine scheduling problems, where the objectives of both problems are to minimize the makespan, and the jobs arrive over time, on two uniform machines with speeds 1 and s （s 〉 1）, and on m identical machines, respectively. For the first problem, the authors show that the on-line LPT algorithm has a competitive ratio of （1 ＋ √5）/2 ≈ 1.6180 and the bound is tight. Furthermore, the authors prove that the on-line LPT algorithm has the best possible competitive ratio if s ≥ 1.8020. For the second problem, the authors present a lower bound of （15 - √17）/8 ≈ 1.3596 on the competitive ratio of any deterministic on-line algorithm. This improves a previous result of 1.3473.展开更多
In this study, we propose and compare stochastic variants of the extra-gradient alternating direction method, named the stochastic extra-gradient alternating direction method with Lagrangian function（SEGL） and the s...In this study, we propose and compare stochastic variants of the extra-gradient alternating direction method, named the stochastic extra-gradient alternating direction method with Lagrangian function（SEGL） and the stochastic extra-gradient alternating direction method with augmented Lagrangian function（SEGAL）, to minimize the graph-guided optimization problems, which are composited with two convex objective functions in large scale.A number of important applications in machine learning follow the graph-guided optimization formulation, such as linear regression, logistic regression, Lasso, structured extensions of Lasso, and structured regularized logistic regression. We conduct experiments on fused logistic regression and graph-guided regularized regression. Experimental results on several genres of datasets demonstrate that the proposed algorithm outperforms other competing algorithms, and SEGAL has better performance than SEGL in practical use.展开更多
基金supported by the Special Funds of the National Natural Science Foundation of China under Grant No.61340045the Specialized Research Fund for the Doctoral Program of Higher Education of China under Grant No.20123705110003Innovation Project of Shangdong Graduate Education under Grant No.SDYC13036
文摘This paper considers two parallel machine scheduling problems, where the objectives of both problems are to minimize the makespan, and the jobs arrive over time, on two uniform machines with speeds 1 and s （s 〉 1）, and on m identical machines, respectively. For the first problem, the authors show that the on-line LPT algorithm has a competitive ratio of （1 ＋ √5）/2 ≈ 1.6180 and the bound is tight. Furthermore, the authors prove that the on-line LPT algorithm has the best possible competitive ratio if s ≥ 1.8020. For the second problem, the authors present a lower bound of （15 - √17）/8 ≈ 1.3596 on the competitive ratio of any deterministic on-line algorithm. This improves a previous result of 1.3473.
基金supported by the National Natural Science Foundation of China(No.61303264)the National Key Research and Development Program of China(No.2016YFB1000401)
文摘In this study, we propose and compare stochastic variants of the extra-gradient alternating direction method, named the stochastic extra-gradient alternating direction method with Lagrangian function（SEGL） and the stochastic extra-gradient alternating direction method with augmented Lagrangian function（SEGAL）, to minimize the graph-guided optimization problems, which are composited with two convex objective functions in large scale.A number of important applications in machine learning follow the graph-guided optimization formulation, such as linear regression, logistic regression, Lasso, structured extensions of Lasso, and structured regularized logistic regression. We conduct experiments on fused logistic regression and graph-guided regularized regression. Experimental results on several genres of datasets demonstrate that the proposed algorithm outperforms other competing algorithms, and SEGAL has better performance than SEGL in practical use.
