摘要
探究学科或领域内研究发展趋势和热点一直以来受到国内外学者们重点关注,而高频关键词的频次变化分析是其中重要的研究内容。关键词的变化与时间存在强相关性,但当前仅有少数研究考虑了关键词随时间密切变化的特性。在考虑关键词信息的时间属性基础上,提出一种基于自适应增强(AdaBoost)的径向基(RBF)神经网络预测算法(以下简称“RBF改进算法”),对关键词频次进行分析预测。对中国知网2007—2022年收录的医学图像期刊论文关键词进行处理,其中将2007年至2021年的数据作为实验训练数据,2022年数据作为验证数据,通过算例分析,对比RBF改进算法、反向传播算法和时间序列算法对关键词词频的预测结果。结果发现:通过AdaBoost算法对RBF算法进行改进,能够增强RBF神经网络的泛化能力以及对样本的适应性,同时保留了RBF神经网络较好的非线性映射能力这一优点;RBF改进算法预测结果与实际数据接近,其预测精度优于反向传播神经网络和时间序列算法,该算法的预测效果更佳。
Exploring research trends and hot topics within a discipline or field has always attracted significant attention from scholars both domestically and internationally,and analyzing the frequency changes of high-frequency keywords is an important aspect of such research.There is a strong correlation between changes in keywords and time,but currently only a few studies have considered the characteristic of keywords changing closely over time.Considering the time attribute of keywords,an AdaBoost-based radial basis function(RBF)neural network forecasting algorithm(hereinafter referred to as"improved RBF algorithm")is proposed and applied to the keyword frequency forecasting and analyzing.The keywords of medical image journal papers indexed in CNKI from 2007 to 2022 are processed,with the data from 2007 to 2021 used as experimental training data and the 2022 data used as validation data.Through case analysis,the prediction results of keyword frequency are compared using the RBF improved algorithm,the backpropagation(BP)algorithm,and auto-regressive moving-average(ARMA)time series analysis algorithm.The results show that improving the RBF algorithm throughthe AdaBoost algorithm can not only enhance the generalization ability and adaptability of the RBF neural network to samples,but also retain the good nonlinear mapping ability of the RBF neural network.The prediction results of the improved RBF algorithm are close to the actual data,and its prediction accuracy is better than the BP neural network algorithm and ARMA time series algorithm.Thus the improved RBF algorithm this paper proposed performs better.
作者
陈张一
朱朝阳
邹玲
胡小君
Chen Zhangyi;Zhu Chaoyang;Zou Ling;Hu Xiaojun(School of Medicine,Zhejiang University,Hangzhou 310058,China)
出处
《科技管理研究》
CSSCI
2024年第18期215-221,共7页
Science and Technology Management Research
基金
国家自然科学基金面上项目“基于‘非典型信号’的变革性研究特征识别与机理辨析的方法论研究及实证考察”(71974169)。
关键词
词频
预测算法
ADABOOST算法
RBF神经网络
算法应用
算法优化
医学图像
the frequency of keywords
forecasting algorithm
AdaBoost algorithm
RBF neural network
algorithm application
algorithm optimization
medical image
