An exploration for higher efficiency in multi objective optimisation with reinforcement learning

Efficiency in optimisation and search processes persists to be one of the challenges, which affects the performance and use of optimisation algorithms. Utilising a pool of operators instead of a single operator to handle move operations within a neighbourhood remains promising, but an optimum or near optimum sequence of operators necessitates further investigation. One of the promising ideas is to generalise experiences and seek how to utilise it. Although numerous works are done around this issue for single objective optimisation, multi-objective cases have not much been touched in this regard. A generalised approach based on multi-objective reinforcement learning approach seems to create remedy for this issue and offer good solutions. This paper overviews a generalisation approach proposed with certain stages completed and phases outstanding that is aimed to help demonstrate the efficiency of using multi-objective reinforcement learning.

Analysing the Predictivity of Features to Characterise the Search Space

Exploring search spaces is one of the most unpredictable challenges that has attracted the interest of researchers for decades. One way to handle unpredictability is to characterise the search spaces and take actions accordingly. A well-characterised search space can assist in mapping the problem states to a set of operators for generating new problem states. In this paper, a landscape analysis-based set of features has been analysed using the most renown machine learning approaches to determine the optimal feature set. However, in order to deal with problem complexity and induce commonality for transferring experience across domains, the selection of the most representative features remains crucial. The proposed approach analyses the predictivity of a set of features in order to determine the best categorization.

A reinforcement learning algorithm for building collaboration in multi-agent systems

This paper presents a proof-of concept study for demonstrating the viability of building collaboration among multiple agents through standard Q learning algorithm embedded in particle swarm optimisation. Collaboration is formulated to be achieved among the agents via some sort competition, where the agents are expected to balance their action in such a way that none of them drifts away of the team and none intervene any fellow neighbours territory. Particles are devised with Q learning algorithm for self training to learn how to act as members of a swarm and how to produce collaborative/collective behaviours. The produced results are supportive to the algorithmic structures suggesting that a substantive collaboration can be build via proposed learning algorithm.

Coordinating metaheuristic agents with swarm intelligence

Coordination of multi agent systems remains as a problem since there is no prominent method to completely solve this problem. Metaheuristic agents are specific implementations of multi-agent systems, which imposes working together to solve optimisation problems with metaheuristic algorithms. The idea borrowed from swarm intelligence seems working much better than those implementations suggested before. This paper reports the performance of swarms of simulated annealing agents collaborating with particle swarm optimization algorithm. The proposed approach is implemented for multidimensional knapsack problem and has resulted much better than some other works published before.