Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

DAG task scheduling method based on Monte Carlo tree search

A task scheduling and tree search technology, applied in other database retrieval, other database indexing, program startup/switching and other directions, can solve problems such as long execution time, and achieve the effect of solving long execution time, improving search efficiency, and ensuring algorithm efficiency

Active Publication Date: 2019-06-07
HANGZHOU DIANZI UNIV
View PDF7 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0013] The present invention aims to overcome the problem of long execution time in the current workflow scheduling method in a distributed environment in the prior art, and provides a method based on Monte Carlo tree search that can accelerate the efficiency of the guaranteed algorithm and improve the search efficiency of the algorithm. DAG task scheduling method

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • DAG task scheduling method based on Monte Carlo tree search
  • DAG task scheduling method based on Monte Carlo tree search
  • DAG task scheduling method based on Monte Carlo tree search

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0043] The present invention will be further described below in conjunction with accompanying drawing and specific embodiment:

[0044] Example: as figure 1 The shown DAG task scheduling method based on Monte Carlo tree search includes the following steps:

[0045] (1-1) Use the CPOP algorithm to find the critical path of the DAG graph;

[0046] (1-2) Selection stage: set the root node of the search tree to S 0 , from the root node S 0 At the beginning, every time a node is passed, it starts to judge whether the node passed through has been expanded;

[0047] If the passed node has not been expanded, enter the expansion stage; if the expansion is completed, select the node with the largest UCT value as the search path node, and the process is calculated using the following formula:

[0048]

[0049]

[0050]

[0051] Among them, Cpuct is an important hyperparameter, which is mainly used to balance the weight between exploration and utilization; N(s,a) represents the...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a DAG task scheduling method based on Monte Carlo tree search. The DAG task scheduling method comprises the following steps: firstly, calculating a critical path of a DAG graphby using a method for solving the critical path in a CPOP algorithm; then executing four stages of Monte Carlo tree search of the method; judging whether the current node is expanded or not from theroot node; if the expansion is finished, selecting the node with the maximum UCT value as a search path node; if the expansion is not finished, adding a new node as an expansion node; starting to simulate a task scheduling process by an expansion node; adopting a random selection strategy to select a processor and a task, obtaining a makspan value after simulation is finished, returning and updating nodes according to the makspan value, and finally a scheduling sequence capable of enabling the makespan value to be minimum is found according to a Monte Carlo tree search result. The method has the characteristic that the search efficiency of the algorithm can be improved while the algorithm efficiency is accelerated and ensured.

Description

technical field [0001] The invention relates to the technical field of task scheduling systems, in particular to a Monte Carlo tree search-based DAG task scheduling method capable of accelerating and ensuring the efficiency of algorithms while improving algorithm search efficiency. Background technique [0002] In distributed heterogeneous computing systems, how to optimize DAG task scheduling is an open research problem. The goal of DAG task scheduling is to provide a scheduling scheme to schedule the tasks in the DAG graph to the processor in a certain order to minimize the scheduling length. Its model is as follows: [0003] A computing application is represented by a directed acyclic graph (DAG) G(V,E). where V represents the set of n tasks in the application, and E represents the set of e edges between tasks. The edge e(i,j)∈E represents the priority constraint, task n j have to wait until n i It can only be executed after it is completed. Usually, a task without ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F9/48G06F16/901G06F16/903
Inventor 程雨夏刘奎吴志伟吴卿
Owner HANGZHOU DIANZI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com