Releases: intLyc/MTO-Platform
Releases · intLyc/MTO-Platform
MToP v1.6
- Fix the bug of multifactorial algorithms run in many-task problems
- New Algorithm:
- TNG-NES (Single-objective Many-task TEVC24)
- MTDE-ADKT (Single-objective Multi-task ASOC24)
- AR-MOEA, MSEA (Multi-objective Single-task)
- New Problem: LSMaTSO (Large-scale many-task single-objective)
MToP v1.5
MToP v1.4
- New features:
- Draw dynamic Dec and Obj of populations during optimization in the Test Module
- Pause and Stop buttons can now respond in time by clicking on both the Test and Experiment Module
- Figures sample numbers in the Test Module can be modified, and figures can be exported
- Algorithm and Problem objects can be input in the command line running e.g. "mto(MFEA(), CMT1());"
- New Algorithms:
- CEDA (Constrained Single-objective Multitask SWEC24)
- MTEA-D-TSD (Multi-objective Multitask GECCO24)
- Global-GA (Single-objective Single-task TEVC24)
- KLDE and KLPSO (Single-objective Single-task TEVC23)
- Other classical algorithms: RVEA (MO-ST), SMS-EMOA (MO-ST), IPOP-CMA-ES (SO-ST)
- New Problems:
- Classical Single-Objective Functions with any dimension setting
- Fix some bugs.
MToP v1.3
MToP v1.2
- Newly added algorithms:
- TRADE (single-objective many-task TCYB 2023)
- ASCMFDE (single-objective multitask TEVC 2021)
- Add error value type of WCCI20-MTSO
- Update Operator GA (SBX and polynomial mutation) with more advanced calculation methods. GA-based algorithms now have improved performance.
MToP v1.1
- The speed of experimental execution is significantly increased, brought by the simultaneous evaluation of whole population decision variables
- 3D task figures of 2-dimensional variables for un-/constrained single-objective multi-/many-/single-task optimization can be plotted in the test module
- Performance metrics can be displayed automatically based on the data type in the experiment module
- Newly added algorithms:
- MKTDE (single-objective multi-task TEVC 2022)
- CCEF-ECHT (constrained single-objective TSMC 2023)
MToP v1.0 for Evolutioanry Multitasking
We introduce the multitask optimization platform, named MToP, for evolutionary multitasking:
- 30+ multitask evolutionary algorithms for multitask optimization
- 30+ single-task evolutionary algorithms can handle multitask optimization problems
- 150+ multitask optimization problem cases with real-world applications
- 10+ performance metrics covering single- and multi-objective optimization
MToP is a user-friendly tool with a graphical user interface that makes it easy to analyze results, export data, and plot schematics. More importantly, MToP is extensible, allowing users to develop new algorithms and define new problems.
MTO-Platform (MToP) v0.36
Add Algorithm: MTES-KG TEVC 2023
Add Problems: Optimal Power Flow as Constrained Multitask Optimization
MTO-Platform (MToP) v0.35
- Update Readme.m
- Add some algorithms of ES
- Modify GUI
- Fix bugs in metrics
MTO-Platform v0.34
- Add Descriptions of Algorithms in "Doc/Algorithms in MTO-Platform"
- Add some algorithms
- Fix some bugs