Research on the Budget of Laboratory Experiment Funds in Colleges and Universities Based on PSO-LS-SVM

In order to improve the accuracy of laboratory experiment budget in universities and better leverage the supporting role of laboratories in teaching and research, this paper combines the actual management of equipment experiment budget in universities. In response to the problems of long training time and slow convergence speed of least squares support vector machine (LS-SVM) under small sample conditions, an improved particle swarm optimization algorithm (IPSO) is introduced to optimize the adjustable parameters and kernel parameters of LS-SVM, and an experimental budget model based on IPSO-LSSVM is constructed. Validate the model through examples and compare it with general methods such as LS-SVM, SVR, and BP neural network. The results show that the method proposed in this paper achieves 0.0352, 0.3225, and 0.9016 in the three evaluation indicators of average absolute percentage error, mean square error, and coefficient of determination, respectively, demonstrating higher accuracy, faster convergence speed, and good generalization ability in small sample budgets.