CAJ | 학술논문

将人工神经网络方法用于DNA分类.首先应用概率统计的方法对20个已知类别的人工DNA序列进行特征提取,形成DNA序列的特征向量,并将之作为样本输入感知器以及BP神经网络进行学习.采用MATLAB软件进行实验,构造了两个神经网络,将提取的DNA特征向量集作为样本分别输入这两个网络进行学习.通过训练后,将20个未分类的人工序列样本和182个自然序列样本提取特征向量并输入两个网络进行分类.结果表明:基于神经网络的分类方法能够以很高的正确率和精度对DNA进行分类,将人工神经网络用于DNA序列分类是完全可行的.
장인공신경망락방법용우DNA분류.수선응용개솔통계적방법대20개이지유별적인공DNA서렬진행특정제취,형성DNA서렬적특정향량,병장지작위양본수입감지기이급BP신경망락진행학습.채용MATLAB연건진행실험,구조료량개신경망락,장제취적DNA특정향량집작위양본분별수입저량개망락진행학습.통과훈련후,장20개미분류적인공서렬양본화182개자연서렬양본제취특정향량병수입량개망락진행분류.결과표명:기우신경망락적분류방법능구이흔고적정학솔화정도대DNA진행분류,장인공신경망락용우DNA서렬분류시완전가행적.
This paper uses the artificial neural network method for DNA typing.First,it use the method of probability and statistics to feature extraction for 20 artificial DNA sequences which categories are known,they form feature vector of DNA sequences,and use the vector as a sample to input sensors and BP neural network for learning.The MATLAB software is used for experiments,two neural network is constructed,extracted characteristic vector set of DNA as a sample input,re-spectively,the two network for learning.Through the training,20 unclassified artificial sequence samples and 182 natural se-quence samples extract the feature vector and enter the two networks for classifying.