This paper proposes a new meta-heuristic optimization algorithm, namely Mud Ring Algorithm (MRA) that mimics the mud ring feeding behaviour of bottlenose dolphins in Atlantic coast of Florida. The inspiration of MRA is mainly based on the foraging behaviour of bottlenose dolphins and their mud ring feeding strategy. This strategy is applied by dolphins to trap fish via creating a plume by a single dolphin moving his tail swiftly in the sand and swims around the group of fish. The fishes become disoriented and jump over the surface only to find the waiting mouths of dolphins. MRA optimization algorithm mathematically simulates this feeding strategy and proves its optimization effectiveness through a comprehensive comparison with other meta-heuristic algorithms. Twenty-nine benchmark functions and four commonly used benchmark engineering challenges are used in the comparison. The statistical comparisons and results prove that the proposed MRA has the superiority in dealing with these optimization problems and can obtain the best solutions than other meta-heuristic optimizers.

Feature selection plays a crucial role in order to mitigate the high dimensional feature space in different classification problems. The computational cost is reduced, and the accuracy of the classification is improved by reducing the dimension of feature space. Hence, in the classification task, finding the optimal subset of features is of utmost importance. Metaheuristic techniques have proved their efficacy in solving many real-world optimization issues. One of the recently introduced physics-inspired optimization methods is Archimedes Optimization Algorithm (AOA). This paper proposes an Enhanced Archimedes Optimization Algorithm (EAOA) by adding a new parameter that depends on the step length of each individual while revising the individual location. The EAOA algorithm is proposed to improve the AOA exploration and exploitation balance and enhance the classification performance for the feature selection issue in real-world data sets. Experiments were performed on twenty-three standard benchmark functions and sixteen real-world data sets to investigate the performance of the proposed EAOA algorithm. The experimental results based on the standard benchmark functions show that the EAOA algorithm provides very competitive results compared to the basic AOA algorithm and five well-known optimization algorithms in terms of improved exploitation, exploration, local optima avoidance and convergence rate. In addition, the results based on sixteen real-world data sets ascertain that reduced feature subset yields higher classification performance when compared with the other feature selection methods.

Instance reduction is a pre-processing step devised to improve the task of classification. Instance reduction algorithms search for a reduced set of instances to mitigate the low computational efficiency and high storage requirements. Hence, finding the optimal subset of instances is of utmost importance. Metaheuristic techniques are used to search for the optimal subset of instances as a potential application. Antlion optimization (ALO) is a recent metaheuristic algorithm that simulates antlion’s foraging performance in finding and attacking ants. However, the ALO algorithm suffers from local optima stagnation and slow convergence speed for some optimization problems. In this study, a new modified antlion optimization (MALO) algorithm is recommended to improve the primary ALO performance by adding a new parameter that depends on the step length of each ant while revising the antlion position. Furthermore, the suggested MALO algorithm is adapted to the challenge of instance reduction to obtain better results in terms of many metrics. The results based on twenty-three benchmark functions at 500 iterations and thirteen benchmark functions at 1000 iterations demonstrate that the proposed MALO algorithm escapes the local optima and provides a better convergence rate as compared to the basic ALO algorithm and some well-known and recent optimization algorithms. In addition, the results based on 15 balanced and imbalanced datasets and 18 oversampled imbalanced datasets show that the instance reduction proposed method can statistically outperform the basic ALO algorithm and has strong competitiveness against other comparative algorithms in terms of four performance measures: Accuracy, Balanced Accuracy (BACC), Geometric mean (G-mean), and Area Under the Curve (AUC) in addition to the run time. MALO algorithm results show increment in Accuracy, BACC, G-mean, and AUC rates up to 7%, 3%, 15%, and 9%, respectively, for some datasets over the basic ALO algorithm while keeping less computational time.

The survival prediction of children undergoing hematopoietic stem-cell transplantation is essential for successful transplantation. However, the performance of current algorithms for predicting mortality in this patient group has not improved over recent decades. This paper proposes a new feature selection technique for survival prediction problems using the Mud Ring Algorithm (MRA). Experiments and tests were initially performed on 13 real datasets with varying occurrences to compare the suggested algorithm with other algorithms. After that, the constructed model classification performance was compared to other techniques using the bone marrow transplant children’s dataset. Modern techniques were used to acquire their classification results, which were then compared to the suggested outcomes using a variety of well-known metrics, graphical tools, and diagnostic analysis. This investigation has demonstrated that our suggested approach is comparable and outperformed other methods in terms of results. In addition, the results showed that the constructed model enhanced prediction accuracy by up to 82.6% for test cases.

Maternal health risk prediction is a critical aspect of public health. This paper proposes a new parameter optimization method to improve maternal health risk prediction using the Mud Ring Algorithm (MRA). The proposed method is implemented in two stages to achieve the desired improvement. In the first stage, the MRA is used to optimize the parameter of Support Vector Machine (SVM) classifier. The MRA-optimized SVM (MRA-SVM) is then evaluated on thirteen real-world datasets to compare its performance with other state-of-the-art optimization algorithms. Subsequently, the performance of the MRA-SVM is specifically compared using a maternal health risk dataset. The maternal health risk dataset, being a medical dataset, faces a significant issue of imbalance. To address this problem, the second stage comes in turn, and the crossover oversampling technique is first employed. Furthermore, other classifiers such as Random Forest and K-Nearest Neighbor are also used for improving the prediction of the maternal health risk dataset. Experimental results indicate that our proposed method delivers highly competitive results compared to six well-known optimization algorithms, evaluated based on Accuracy, G-mean, F-Measure, MCC and Kappa metrics. Moreover, the results demonstrate that using a crossover oversampling method as a preprocessing step, along with MRA to optimize the parameters of SVM (MRA-SVM), Random Forest (MRA-RF), and K-Nearest Neighbor (MRA-KNN), increases prediction accuracy of maternal health risk dataset by 11.8%, 9.11%, and 17.08% respectively, compared to using the three classifiers without MRA and crossover oversampling. In sum, utilizing the crossover oversampling method, combined with the MRA to optimize the parameter of Random Forest classifier, leads to higher prediction performance compared to other recently published algorithms.