AWD-stacking: An enhanced ensemble learning model for predicting glucose levels.

Accurate prediction of blood glucose levels is essential for type 1 diabetes optimizing insulin therapy and minimizing complications in patients with type 1 diabetes. Using ensemble learning algorithms is a promising approach. In this regard, this study proposes an improved stacking ensemble learning algorithm for predicting blood glucose level, in which three improved long short-term memory network models are used as the base model, and an improved nearest neighbor propagation clustering algorithm is adaptively weighted to this ensemble model. The OhioT1DM dataset is used to train and evaluate the performance of the proposed model. This study evaluated the performance of the proposed model using the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Matthews Correlation Coefficient (MCC) as the evaluation metrics. The experimental results demonstrate that the proposed model achieves an RMSE of 1.425 mg/dL, MAE of 0.721 mg/dL, and MCC of 0.982 mg/dL for a 30-minute prediction horizon(PH), RMSE of 3.212 mg/dL, MAE of 1.605 mg/dL, and MCC of 0.950 mg/dL for a 45-minute PH; and RMSE of 6.346 mg/dL, MAE of 3.232 mg/dL, and MCC of 0.930 mg/dL for a 60-minute PH. Compared with the best non-ensemble model StackLSTM, the RMSE and MAE were improved by up to 27.92% and 65.32%, respectively. Clarke Error Grid Analysis and critical difference diagram revealed that the model errors were within 10%. The model proposed in this study exhibits state-of-the-art predictive performance, making it suitable for clinical decision-making and of significant importance for the effective treatment of diabetes in patients.

Improved intelligent clonal optimizer based on adaptive parameter strategy.

The intelligent clonal optimizer (ICO) is a new evolutionary algorithm, which adopts a new cloning and selection mechanism. In order to improve the performance of the algorithm, quasi-opposition-based and quasi-reflection-based learning strategy is applied according to the transition information from exploration to exploitation of ICO to speed up the convergence speed of ICO and enhance the diversity of the population. Furthermore, to avoid the stagnation of the optimal value update, an adaptive parameter method is designed. When the update of the optimal value falls into stagnation, it can adjust the parameter of controlling the exploration and exploitation in ICO to enhance the convergence rate of ICO and accuracy of the solution. At last, an improved intelligent chaotic clonal optimizer (IICO) based on adaptive parameter strategy is proposed. In this paper, twenty-seven benchmark functions, eight CEC 2104 test functions and three engineering optimization problems are used to verify the numerical optimization ability of IICO. Results of the proposed IICO are compared to ten similar meta-heuristic algorithms. The obtained results confirmed that the IICO exhibits competitive performance in convergence rate and accurate convergence.