An Improved BPNN Prediction Method Based on Multi-Strategy Sparrow Search Algorithm

Abstract

Data prediction can improve the science of decision-making by making predictions about what happens in daily life based on natural law trends. Back propagation (BP) neural network is a widely used prediction method. To reduce its probability of falling into local optimum and improve the prediction accuracy, we propose an improved BP neural network prediction method based on a multi-strategy sparrow search algorithm (MSSA). The weights and thresholds of the BP neural network are optimized using the sparrow search algorithm (SSA). Three strategies are designed to improve the SSA to enhance its optimization-seeking ability, leading to the MSSA-BP prediction model. The MSSA algorithm was tested with nine different types of benchmark functions to verify the optimization performance of the algorithm. Two different datasets were selected for comparison experiments on three groups of models. Under the same conditions, the mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) of the prediction results of MSSA-BP were significantly reduced, and the convergence speed was significantly improved. MSSA-BP can effectively improve the prediction accuracy and has certain application value.

Improving the Accuracy of TanDEM-X Digital Elevation Model Using Least Squares Collocation Method

“…Table 2 shows the performance comparison between the LSC-TXC algorithm proposed in this paper and various other models for DEM correction. It can be found that the ME and RMSE of TanDEM-X DEM corrected by LSC-TXC, Random Forest (RF) [42], Height Difference Fitting Neural Network (HDFNN) [43], and Back Propagation Neural Network (BPNN) [44] are reduced compared to the TanDEM-X DEM (ME = −2.019 m, RMSE = 6.141 m), indicating that the four correction models have effectively enhanced the accuracy of the original TanDEM-X DEM in this study area. The ME of the RF (−0.033 m) outperforms that of the LSC-TXC algorithm proposed in this paper (0.058 m).…”

Multi-Objective Gannet Optimization Algorithm for Dynamic Passenger Flow Allocation in Train Operation Plan Optimization

“…In the early stage, the number of individuals in the constrained search population is relatively large, but with the iteration of the algorithm, the number of individuals N 1 in the population is dynamically reduced, as shown in Eq. (36). The second population that performs unconstrained search has a small number of individuals in the early stage, but as the algorithm is iteratively updated, the number of offspring individuals N 2 is constantly increasing, as shown in Eq.…”

“…Section IV introduces the design rules of MOGOA in detail, and the adopted improved strategy of adaptive multi-population co-evolution. Section V, we compare MOGOA with three outstanding multiobjective optimization algorithms: NSGA-II [34], Multi-Objective Gray Wolf Optimizer (MOGWO) [35], and Multistrategy Sparrow Search Algorithm (MSSA) [36] in several test suites of ZDT1-4, 6 and WFG1-5 or the problem of driving organization optimization, and give an analysis of the experimental results. Finally, we conclude the entire paper in Section VI.…”

UAV mountain material transportation based on an improved sparrow search algorithm

“…Ge et al [18] enhanced the diversity of the initial population by using Sobol sequences and used nonlinear inertia weights to control the search range and exploration efficiency. Zhang et al [19] optimized the initial population using Circle chaotic mapping, integrated the Butterfly Optimization Algorithm to enhance the algorithm's global exploitation capability, and employed a dimension-by-dimension mutation strategy to help the algorithm escape from local optima.…”