Scheduling seminar

Download

schedulingseminar@rtime.felk.cvut.cz

October 2021

1 participants
2 discussions

Benjamin Moseley (Carnegie Mellon) | October 27 | Machine Learning for Scheduling
by Zdeněk Hanzálek 25 Oct '21

25 Oct '21

Dear scheduling researcher, We are delighted to announce the talk given by Benjamin Moseley <http://www.andrew.cmu.edu/user/moseleyb/> (Carnegie Mellon). The title is "Machine Learning for Scheduling". The seminar will take place on Zoom on Wednesday, October 27 at 13:00 UTC. Join Zoom Meeting https://cesnet.zoom.us/j/96553570741?pwd=ZkcwUk1IZjM1R2xDSkNQUWtYOHNVdz09 <https://cesnet.zoom.us/j/96553570741?pwd=ZkcwUk1IZjM1R2xDSkNQUWtYOHNVdz09> Meeting ID: 965 5357 0741 Passcode: 705618 You can follow the seminar online or offline on our Youtube channel as well: https://www.youtube.com/channel/UCUoCNnaAfw5NAntItILFn4A The abstract follows. This talk will discuss a model for augmenting algorithms with useful predictions that go beyond worst-case bounds on the algorithm performance. The model ensures predictions are formally learnable and instance robust. Learnability guarantees that predictions can be efficiently constructed from past data. Instance robustness formally ensures a prediction is robust to modest changes in the problem input. This talk will discuss predictions that satisfy these properties for scheduling and resource augmentation. Algorithms developed break through worst-case barriers with accurate predictions and have a graceful degradation in performance when the error in the predictions grows. The next talk in our series will be given by Carlo Mannino (SINTEF & Oslo Uni.) | November 10 | Train Scheduling: Models, decomposition methods and practice. For more details, please visit https://schedulingseminar.com/ With kind regards Zdenek, Mike and Guohua -- Zdenek Hanzalek Industrial Informatics Department, Czech Institute of Informatics, Robotics and Cybernetics, Czech Technical University in Prague, Jugoslavskych partyzanu 1580/3, 160 00 Prague 6, Czech Republic https://rtime.ciirc.cvut.cz/~hanzalek/

1 0

Federico Della Croce (DIGEP - Polito.it) | October 13 | The Longest Processing Time Rule for Identical Parallel Machines Revisited
by Zdeněk Hanzálek 11 Oct '21

11 Oct '21

Dear scheduling researcher, We are delighted to announce the talk given by Federico Della Croce (DIGEP - Polito.it). The title is "The Longest Processing Time Rule for Identical Parallel Machines Revisited". The seminar will take place on Zoom on Wednesday, October 13 at 13:00 UTC. Join Zoom Meeting https://cesnet.zoom.us/j/95567469894?pwd=cDcyZGRWaTVRTVptOEUvNlIrTUpOZz09 <https://cesnet.zoom.us/j/95567469894?pwd=cDcyZGRWaTVRTVptOEUvNlIrTUpOZz09> Meeting ID: 955 6746 9894 Passcode: 138314 You can follow the seminar online or offline on our Youtube channel as well: https://www.youtube.com/channel/UCUoCNnaAfw5NAntItILFn4A The abstract follows. We consider the P||Cmax scheduling problem where the goal is to schedule n jobs on m identical parallel machines to minimize makespan. We revisit the famous Longest Processing Time (LPT) rule proposed by Graham in 1969. LPT requires sorting jobs in non-ascending order of processing times and then assigning one job at a time to the machine whose load is smallest so far. We provide new insights into LPT and discuss the approximation ratio of a modification of LPT that improves Graham’s bound. We use linear programming to analyze the approximation ratio of our approach. This performance analysis can be seen as a valid alternative to formal proofs based on analytical derivation. Also, we derive from the proposed approach an O(n log n) time complexity heuristic called SLACK. The heuristic splits the sorted job set in tuples of m consecutive jobs (1,...,m; m+1,...,2m; etc.) and sorts the tuples in non-increasing order of the difference (SLACK) between the largest and smallest job in the tuple. Given this new ordering of the job set, list scheduling is applied. This approach strongly outperforms LPT on benchmark literature instances and is competitive with more involved approaches such as COMBINE and LDM. The next talk in our series will be given by Benjamin Moseley <http://www.andrew.cmu.edu/user/moseleyb/> (Carnegie Mellon) | October 27 | Machine Learning for Scheduling. For more details, please visit https://schedulingseminar.com/ With kind regards Zdenek, Mike and Guohua -- Zdenek Hanzalek Industrial Informatics Department, Czech Institute of Informatics, Robotics and Cybernetics, Czech Technical University in Prague, Jugoslavskych partyzanu 1580/3, 160 00 Prague 6, Czech Republic https://rtime.ciirc.cvut.cz/~hanzalek/

1 0