Source language parse trees offer very useful but imperfect reordering constraints for statistical machine translation. A lot of effort has been made for soft applications of syntactic constraints. We alternatively propose the selective use of syntactic constraints. A classifier is built automatically to decide whether a node in the parse trees should be used as a reordering constraint or not. ...

We investigate the problems of scheduling n jobs tom = m1+ m2 identical machines where m1 machines are always available, m2 machines have some specified unavailable intervals.The objective is tominimize the makespan. We assume that if a job is interrupted by the unavailable interval, it can be resumed after the machine becomes available. We show that if at least one machine is always available,...

This paper deals with a three-machine flow-shop problem in which some of the job sequences are infeasible. It is assumed that jobs are grouped into several disJoint subsets within which a job order is pre-determined. Once the first job in a group has started on a machine, then the entire group must be completed on the machine without starting a job which does not belong to the group. It is furt...

We consider parallel, identical machine scheduling problems where the jobs are subject to precedence constraints, release dates, and the processing times of jobs are governed by independent probability distributions. The objective is to minimize the expected value of the total weighted completion time ∑w j C j , where w j 0. Building upon a linear programming relaxation by Möhring, Schulz, and ...

Suppose jobs, in the Johnson's two-machine n-job flow-shop scheduling problem, are grouped into disjoint subsets within which a job order that may not be preempted is specified. Furthermore, suppose that a precedence relation between these subsets is given such that the processing of a subset must be completed, on each machine, before the processing of another subset begins on the machine. This...

This paper presents experiments with document-level machine translation with readability constraints. We describe the task of producing simplified translations from a given source with the aim to optimize machine translation for specific target users such as language learners. In our approach, we introduce global features that are known to affect readability into a documentlevel SMT decoding fr...

Distributed Machine Learning with Communication Constraints by Yuchen Zhang Doctor of Philosophy in Computer Science University of California, Berkeley Professor Michael I. Jordan, Co-chair Professor Martin J. Wainwright, Co-chair Distributed machine learning bridges the traditional fields of distributed systems and machine learning, nurturing a rich family of research problems. Classical machi...