Multi-Task Multi-Behavior MAP-Elites

05/02/2023
by   Mouret, et al.
0

We propose Multi-Task Multi-Behavior MAP-Elites, a variant of MAP-Elites that finds a large number of high-quality solutions for a large set of tasks (optimization problems from a given family). It combines the original MAP-Elites for the search for diversity and Multi-Task MAP-Elites for leveraging similarity between tasks. It performs better than three baselines on a humanoid fault-recovery set of tasks, solving more tasks and finding twice as many solutions per solved task.

READ FULL TEXT
research
03/09/2020

Quality Diversity for Multi-task Optimization

Quality Diversity (QD) algorithms are a recent family of optimization al...
research
04/23/2022

Exploring the Role of Task Transferability in Large-Scale Multi-Task Learning

Recent work has found that multi-task training with a large number of di...
research
07/11/2021

Self-Referential Quality Diversity Through Differential Map-Elites

Differential MAP-Elites is a novel algorithm that combines the illuminat...
research
05/24/2018

Basket Completion with Multi-task Determinantal Point Processes

Determinantal point processes (DPPs) have received significant attention...
research
01/23/2019

Enhancing MapReduce Fault Recovery Through Binocular Speculation

MapReduce speculation plays an important role in finding potential task ...
research
08/23/2011

Ant Colony Optimization of Rough Set for HV Bushings Fault Detection

Most transformer failures are attributed to bushings failures. Hence it ...
research
09/23/2022

Do Current Multi-Task Optimization Methods in Deep Learning Even Help?

Recent research has proposed a series of specialized optimization algori...

Please sign up or login with your details

Forgot password? Click here to reset