摘要
为了提高疾病诊断的准确率,提出了粒子群算法优化极限学习机的疾病诊断方法。传统的极限学习机算法初始权值和阈值随机产生,限制了算法的预测精度。为了改进极限学习机的预测效果,将极限学习机的权值和阈值作为粒子群中粒子的位置,通过粒子群的全局寻优获得期望的权值和阈值,提升极限学习机的神经元敏感度。将优化后的极限学习机算法与传统预测算法进行仿真对比,结果验证了改进极限学习机的有效性。在疾病诊断上,相比于其他传统算法,所设计的算法具有更高的预测精度和收敛速度,验证了所提方法的可靠性。
Particle swarm optimization(PSO) is proposed to optimize the extreme learning machine(PSO-ELM) in order to improve the accuracy of disease diagnosis. The traditional extreme learning machine algorithm produces initial weights and thresholds randomly, which limits the prediction accuracy of the ELM. In order to improve the prediction accuracy of the ELM, the weights and thresholds of the ELM are taken as the positions of particles in the PSO, and the desired weights and thresholds are obtained through the global optimization of particle swarm, so as to improve the neuronal sensitivity of the ELM. Simulation results show the effectiveness of the improved extreme learning machine by comparing the optimized extreme learning machine algorithm with the traditional predictive algorithm. Compared with other traditional algorithms, the proposed algorithm has higher prediction accuracy and convergence speed, which verifies the reliability of the proposed method in disease diagnosis.
作者
张杜娟
王震
Zhang Dujuan;Wang Zhen(School of Health Services Management,Xi’an Medical University,Xi’an 710021,China;School of Science,Xijing University,Xi’an 710123,China)
出处
《国外电子测量技术》
北大核心
2021年第8期82-86,共5页
Foreign Electronic Measurement Technology
基金
陕西省教育厅专项科研计划(19JK0770)
陕西省自然科学基础研究计划(2020JM-646)项目资助。
关键词
极限学习机
粒子群
疾病诊断
extreme learning machine
particle swarm
disease diagnosis
