• Corpus ID: 17722510

Multi-Objective Deep Q-Learning with Subsumption Architecture

  title={Multi-Objective Deep Q-Learning with Subsumption Architecture},
  author={Tomasz Tajmajer},
  • T. Tajmajer
  • Published 21 April 2017
  • Computer Science
  • ArXiv
In this work we present a method for using Deep Q-Networks (DQNs) in multi-objective tasks. Deep Q-Networks provide remarkable performance in single objective tasks learning from high-level visual perception. However, in many scenarios (e.g in robotics), the agent needs to pursue multiple objectives simultaneously. We propose an architecture in which separate DQNs are used to control the agent's behaviour with respect to particular objectives. In this architecture we use signal suppression… 

Figures and Tables from this paper

A Multi-Objective Deep Reinforcement Learning Framework
  • T. Nguyen
  • Computer Science
    Eng. Appl. Artif. Intell.
  • 2020
Combining a gradient-based method and an evolution strategy for multi-objective reinforcement learning
A two-stage MORL framework combining a gradient-based method and an evolution strategy is proposed to learn multiple policies collaboratively, and experimental results show the superiority of the proposed method.
ETM: Effective Tuning Method Based on Multi-Objective and Knowledge Transfer in Image Recognition
An effective tuning method based on multi-objective and knowledge transfer, which is solved the above limitations in the image recognition and improves the efficiency of the above tuning process by transferring knowledge.
Multi-objective multi-agent decision making: a utility-based analysis and survey
A new taxonomy is developed which classifies multi-objective multi-agent decision making settings, on the basis of the reward structures, and which and how utility functions are applied, and defines and discusses these solution concepts under both ESR and SER optimisation criteria.
Review: Generic MODRL
The use of linear and non-linear methods are used to develop the framework which is able to accommodate both single-policy and multi-policy strategies and indicate the convergence to the optimal Pareto solutions effectively.


Multi-Objective Deep Reinforcement Learning
We propose Deep Optimistic Linear Support Learning (DOL) to solve high-dimensional multi-objective decision problems where the relative importances of the objectives are not known a priori. Using
Human-level control through deep reinforcement learning
This work bridges the divide between high-dimensional sensory inputs and actions, resulting in the first artificial agent that is capable of learning to excel at a diverse array of challenging tasks.
Multiple-Goal Reinforcement Learning with Modular Sarsa(0)
A new algorithm, GM-Sarsa(O), for finding approximate solutions to multiple-goal reinforcement learning problems that are modeled as composite Markov decision processes, which finds good policies in the context of the composite task.
Playing Atari with Deep Reinforcement Learning
This work presents the first deep learning model to successfully learn control policies directly from high-dimensional sensory input using reinforcement learning, which outperforms all previous approaches on six of the games and surpasses a human expert on three of them.
Multi-objective reinforcement learning using sets of pareto dominating policies
A novel temporal difference learning algorithm that integrates the Pareto dominance relation into a reinforcement learning approach and outperforms current state-of-the-art MORL algorithms with respect to the hypervolume of the obtained policies.
Multi-objectivization of reinforcement learning problems by reward shaping
It is shown that adding several correlated signals can help to solve the basic single objective problem faster and better, and it is proved that the total ordering of solutions, and by consequence the optimality of solutions is preserved.
Behavior hierarchy learning in a behavior-based system using reinforcement learning
Hand-design of an intelligent agent's behaviors and their hierarchy is a very hard task. One of the most important steps toward creating intelligent agents is providing them with capability to learn
Scalarized multi-objective reinforcement learning: Novel design techniques
The Chebyshev scalarization method overcomes the flaws of the linear scalarized function as it can discover Pareto optimal solutions regardless of the shape of the front, i.e. convex as well as non-convex.
Multiobjective Reinforcement Learning: A Comprehensive Overview
The basic architecture, research topics, and naïve solutions of MORL are introduced at first and several representative MORL approaches and some important directions of recent research are comprehensively reviewed.