1: \begin{abstract}

2: We introduce FEDQ-Trust, an innovative data-driven trust prediction approach designed for mobile edge-based Internet of Things (IoT) environments. The decentralized nature of mobile edge environments introduces challenges due to variations in data distribution, impacting the accuracy and training efficiency of existing distributed data-driven trust prediction models.

3: % \textbf{In MEC-base IoT system, trust in IoT devices critically affects system security and data integrity. However, the distributed nature of MEC results in significant differences in data distribution. These statistical heterogeneity issues challenge the accuracy and training efficiency of trust prediction models.}

4: % FEDQ-Trust combines \underline{F}ederated \underline{E}xpectation-Maximization  with \underline{D}eep \underline{Q} Networks to address this issue. 

5: FEDQ-Trust effectively tackles the statistical heterogeneity challenges by integrating \underline{F}ederated \underline{E}xpectation-Maximization  with \underline{D}eep \underline{Q} Networks.

6: Federated Expectation-Maximization's robust handling of statistical heterogeneity significantly enhances trust prediction accuracy.

7: % In this regard, it addresses the challenge of varied data distributions across mobile edge nodes. 

8: Meanwhile, Deep Q Networks streamlines the model training process, efficiently reducing the number of training clients while maintaining model performance.

9: % FEDQ-Trust stands as a comprehensive solution, effectively balancing the intricacies of accuracy in handling diverse data with the need for efficiency in model training. 

10: We conducted a suite of experiments within simulated MEC-based IoT settings by leveraging two real-world IoT datasets. 

11: The experimental results demonstrate that our model achieved a significant convergence time reduction of 97\% to 99\% while ensuring a notable improvement of 8\% to 14\% in accuracy compared to state-of-the-art models. 

12: \end{abstract}

13: