CAJ | 학술논문

文章根据大脑信息处理原理，对模块化神经网络进行相关的解析，分析模块化神经网络的信息点，对于由大量信息点组成的子网络结构学习自组织问题，神经元实现对输入样本空间的链接等问题提出了相关的算法，解决了子网络在有效时间段内采集相关数据，并实现集成的问题，提高了神经网络学习的效率，也对获取学习信息进行了相关的研究。
문장근거대뇌신식처리원리，대모괴화신경망락진행상관적해석，분석모괴화신경망락적신식점，대우유대량신식점조성적자망락결구학습자조직문제，신경원실현대수입양본공간적련접등문제제출료상관적산법，해결료자망락재유효시간단내채집상관수거，병실현집성적문제，제고료신경망락학습적효솔，야대획취학습신식진행료상관적연구。
Based on the treatment principle for brain information, the paper analyses the modularized neural network and its information bit, and proposes the relevant algorithm in accordance with self-organiszation learning problemfor a sub network structure constituting of large amount of information bit, and with realizing the links about neurons to input sample space.So, the paper solves the data gath-ering in effective period for sub network, realizes the integrated effect, improves the efficiency of neural network learning, also has car-ried on the related study on access to learning information.