Learning Algorithm And Application Research Based On The Deep Convolutional Neural Network

Deep learning is a new field in the process of machine learning research, AS a neural network it's the purpose is to simulate the human brain to learn,And it can mimic the operation of the brain to process related data,it makes the machine can be like humans,Have strong analysis ability to learn,This is the deep learning's goals.The machines can identify voice, image and text data effectively by deep learning.The essence of deep learning is to learn more useful feature by building machine learning models which has a lot of hidden layers and buliding a large number of training data, In order to learn more useful features,Thus to improve the accuracy of the classification or prediction.Convolutional Neural Network as a machine learning model under the depth supervision learning,It has the characteristics of high learning efficiency.and become one of the highlights in agro-scientific research in the many fields of science,especially the field of pattern classification has achieved high grades.In this paper, through the process of learning and combed the convolutional neural network both at home and abroad and the latest research results,Will be detailed introduces the structure and characteristics of convolution neural network model,and according to the convolution neural network is more suitable for large data sets, and puts forward an improved algorithm for small sample,bring in the dropout,and in terms of handwritten Chinese characters image application, so that achive good results; besides, Also use the convolutional neural network application in biological DNA fingerprint recognition algorithm, by improving the biological DNA fingerprint recognition method, So as to improve the accuracy of biological DNA fingerprint identification.