Skip to content

Latest commit

 

History

History
147 lines (100 loc) · 4.12 KB

User-Guide.md

File metadata and controls

147 lines (100 loc) · 4.12 KB

🧭 MTO Platform User Guide

→ 中文指南

Quick Start

Run MTO Platform

  • GUI: mto
  • Command-Line Examples:
mto({MFEA, MFDE},{CMT1, CMT2})
mto({MFEA, MFDE},{CMT1, CMT2}, 5, true, 100, false, 'MTOData.mat', 2333)
mto({MFEA, MFDE},{CMT1, CMT2}, 'Reps', 5, 'Par_Flag', true)

Add Your Algorithm

  1. Inherit the Algorithm.m class in the Algorithms folder.
  2. Implement the method:
function run(Algo, Prob)
  1. Add labels in the second line:
<Multi-task/Many-task/Single-task> <Multi-objective/Single-objective> <None/Competitive/Constrained>
  1. Refer to MFEA or MO-MFEA implementations for guidance.

Add Your Problem

  1. Inherit the Problem.m class in the Problem folder.
  2. Implement the method:
function Tasks = setTasks()
  1. Add labels in the second line:
<Multi-task/Many-task/Single-task> <Multi-objective/Single-objective> <None/Competitive/Constrained>
  1. Refer to CEC17_MTSO or MTMO_Instance1 for examples.

Add Your Metric

  1. Inherit the Metric.m class in the Metric folder.
  2. Add labels in the second line: <Metric>
  3. Refer to Obj.m and IGD.m implementations.

Modules

1. Test Module

  • Algorithm selection: select and open an algorithm to modify parameters (double-click).

  • Problem selection: select and open a problem to modify parameters.

  • Run: click Start.

  • View figures:

    • 1D Task Figure (normalized / real)
    • 2D Task Figure (normalized / real)
    • 2D Feasible Region
    • Convergence plot
    • Pareto Front

2. Experiment Module

  1. Parameters:

    • Repetitions: number of independent runs
    • Data Length: convergence data length
    • Save Dec: save decision variables flag
    • Parallel: parallel execution flag

  2. Algorithm selection:

    • Click Add to add algorithms to Selected Algorithms
    • Expand and double-click to edit parameters or algorithm names
    • Multi-selectable; right-click to select all

  3. Problem selection:

    • Click Add to add problems to Selected Problems
    • Expand and double-click to edit parameters or problem names
    • Multi-selectable; right-click to select all

  4. Start / Pause / Stop

    • Click Start to run experiments
    • Pause and Resume to control running
    • Stop to terminate

  5. Table Statistics:

    • Display metrics and draw convergence plots / Pareto front
    • Data type: Mean, Mean&Std, Std, Median, Best, Worst
    • Statistical test: None, Wilcoxon, Friedman
    • Highlight data: None, Highlight best, Highlight best & worst
    • Click Save to save table data

  6. Read / Save Data

    • Click Save Data to store experiment results
    • Click Load Data to read saved data

3. Data Process Module

  1. Read Data: click Load Data to add data to Data Tree
  2. Delete Data: select and click Delete Data
  3. Save Data: select and click Save Data
  4. Data Split: by Reps, Algorithms, Problems
  5. Data Merge: by Reps, Algorithms, Problems

Tips

  • Multi-task / many-task problems can be handled via selected algorithms
  • Parallel flag is only effective for independent repetitions
  • Always save your data after Metric calculations