Have a personal or library account? Click to login
Tackling Uncertainty in Reinforcement Learning: A Dual Variational Inference Approach for Task and State Estimation Cover

Tackling Uncertainty in Reinforcement Learning: A Dual Variational Inference Approach for Task and State Estimation

Open Access
|Sep 2025

Figures & Tables

Figure 1.

Gray Uncertainty Information and Fuzzy Uncertainty Information
Gray Uncertainty Information and Fuzzy Uncertainty Information

Figure 2.

Task Inference Network
Task Inference Network

Figure 3.

State Inference Network
State Inference Network

Figure 4.

Decision Network Structure
Decision Network Structure

Figure 5.

Model Network Architecture
Model Network Architecture

Figure 6.

Algorithm Comparison Experiment Diagram Under Different Environments
Algorithm Comparison Experiment Diagram Under Different Environments

Figure 7.

Navigation Comparison Experiment
Navigation Comparison Experiment

Figure 8.

Ablation Experiment
Ablation Experiment

Figure 9.

Noise Resistance Experiment
Noise Resistance Experiment

Ablation Test Results

ModelConvergence StepsAverage Reward
DTS-Infer400000+-21.45
DTS-Infer(with out priori)1000000+-37.36
DTS-Infer(with out zc)1000000+-62.66
TD3(with out zc,zs)1000000+-113.35

Comparative Test Results

EnvironmentModelConvergence StepsAverage Reward
Half-Cheetah-VelDTS-Infer400000+-21.45
PEARL600000+-35.76
Half-Cheetah-Fwd-BackDTS-Infer400000+1612.61
PEARL1000000+1356.40
Ant-GoalDTS-Infer1000000+-193.36
PEARL1000000+-292.69
Language: English
Page range: 101 - 115
Published on: Sep 30, 2025
Published by: Xi’an Technological University
In partnership with: Paradigm Publishing Services
Publication frequency: 4 issues per year

© 2025 Zhidong Yang, Haoyu Liu, Zongxin Yao, Hongge Yao, published by Xi’an Technological University
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.