科技通报
科技通報
과기통보
BULLETIN OF SCIENCE AND TECHNOLOGY
2013年
12期
1-3
,共3页
非稳态波动%分类模型%N平均-依赖过滤%偏最小二乘
非穩態波動%分類模型%N平均-依賴過濾%偏最小二乘
비은태파동%분류모형%N평균-의뢰과려%편최소이승
unsteady fluctuation%classification model%N-dependent on average filter%partial under the squares
提出了一种基于N平均-依赖过滤和偏最小二乘方法的分类模型，基于数据序列的当前条件方差塑造数学非稳态波动模型，采用N平均-依赖数据过滤模型获取N个-依赖后验概率的均值当成分类概率，对数学非稳态波动模型进行初步分类，采用偏最小二乘方法对非稳态波动模型进行再次分类，增强模型的分类精度。实验结果说明，该种方法具有较高的抗干扰性，对非稳态波动数据分类的效率和精度都优于传统模型，具有较高的应用价值。
提齣瞭一種基于N平均-依賴過濾和偏最小二乘方法的分類模型，基于數據序列的噹前條件方差塑造數學非穩態波動模型，採用N平均-依賴數據過濾模型穫取N箇-依賴後驗概率的均值噹成分類概率，對數學非穩態波動模型進行初步分類，採用偏最小二乘方法對非穩態波動模型進行再次分類，增彊模型的分類精度。實驗結果說明，該種方法具有較高的抗榦擾性，對非穩態波動數據分類的效率和精度都優于傳統模型，具有較高的應用價值。
제출료일충기우N평균-의뢰과려화편최소이승방법적분류모형，기우수거서렬적당전조건방차소조수학비은태파동모형，채용N평균-의뢰수거과려모형획취N개-의뢰후험개솔적균치당성분류개솔，대수학비은태파동모형진행초보분류，채용편최소이승방법대비은태파동모형진행재차분류，증강모형적분류정도。실험결과설명，해충방법구유교고적항간우성，대비은태파동수거분류적효솔화정도도우우전통모형，구유교고적응용개치。
Put forward a kind of based on the average-dependent on N filtering and classification model of partial least squares method based on the current condition of variance of the data sequence model unsteady wave model, the average N-dependent data filtering model for N-rely on the a posteriori probability as average classification probability, a pre-liminary classification of unsteady fluctuation of mathematics model, by using partial least square method for unsteady fluctuation model again classified, enhance the classification accuracy of the model. Experimental results indicate that the method has high anti-interference, the unsteady fluctuation data classification is better than the traditional model, the ef-ficiency and accuracy of has higher application value.
