Improved LightGBM Algorithm Based on Swarm Intelligence Optimization and Application
A prediction algorithm based on swarm intelligence optimization is proposed to address the challenge of complex parameter tuning in the LightGBM(Light Gradient Boosting Machine)algorithm.This algorithm adopts multiple swarm in-telligence optimization methods to comprehensively optimize the key hyperparameters of the LightGBM model,including the maximum depth of the tree model,the maximum number of leaf nodes in each tree,the minimum number of samples required for leaf nodes,sub sample ratio,feature sub sampling ratio,and learning rate.To verify the effectiveness of the algorithm,experiments were conducted on experimental data for heart failure disease classification.The experimental results showed that the improved LightGBM heart failure classification model showed significant advantages in key indicators such as stability and accuracy,outperforming K-nearest neighbor algorithm,decision tree,grid search optimization based LightGBM,and random search optimization based LightGBM algorithm.
LightGBM modelswarm intelligence optimizationmodel selectionclassification of heart failure diseases
NETL
NSTL
万方数据
