摘要
频率分布序列预测是成分数据研究中一类重要的问题。融合频率分布所具有的统计规律性与优化思想。文章采用灰色预测理论与最优化方法相结合的预测方法,以期望与均值的预测值之间差异最小化为目标,设置历史经验约束与成分数据约束的组合为约束条件,构建二次规划数学预测模型,并通过某专业学生专业核心能力频率分布结构数据进行实证分析,验证了所提预测模型的有效性。结果表明,所提预测模型的平均绝对误差为0.0416,均方根误差为0.0460,方向余弦为0.9824,与球坐标变换预测结果相比具有更好的预测精度,预测模型是可行的。研究表明,所提预测模型具有较好的预测精度,可以有效解决频率分布序列预测问题,同时对成分数据的分量中含有0或1的预测问题提供了一个新的研究思路。
Frequency distribution series prediction is an important class of problems in compositional data research.By integrating the statistical regularity of frequency distribution and optimization idea,the prediction method combining grey prediction theory and optimization method are used to construct the mathematical prediction model for quadratic programming by taking the minimization of the difference between the predicted values of expectation and mean as the goal,and setting the combination of historical experience constraints and constituent data constraints as the constraints.The empirical analysis of the frequency distribution structure data of core competencies among students in a certain major are used to verify the the effectiveness of the proposed prediction model.The results show that the average absolute error of the proposed prediction model is 0.0416,the root mean square error is 0.0460,and the directional cosine is 0.9824.In comparison with the spherical coordinate transformation prediction results,the proposed model has better prediction accuracy and is feasible.The research results show that the proposed prediction model has better prediction accuracy and can effectively solve the problem of frequency distribution sequence prediction,and meanwhile it can provide a new research idea for the prediction problem of compositional data containing 0 or 1 in the compositional data.
作者
喻芳宇
高胜哲
YU Fangyu;GAO Shengzhe(College of Information Engineering,Dalian Ocean University,Dalian 116023,China;MOE Key Laboratory of Environment Controlled Aquaculture,Dalian Ocean University,Dalian 116023,China)
出处
《现代电子技术》
北大核心
2024年第2期171-175,共5页
Modern Electronics Technique
基金
辽宁省教育厅基本科研项目(面上项目)(LJKZ0732)
设施渔业教育部重点实验室资助(202321)。
关键词
成分数据
随机统计
频率分布
灰色预测模型
数学预测模型
专业核心能力
compositional data
random statistics
frequency distribution
grey prediction model
mathematical prediction models
professional core competencies
