Design And Development Of Image Recognition And Sentence Recommendation System Based On Deep Learning

Deep leaming(DL)is an important method and research direction of machine learning(ML),which is an important branch of Artificial Intelligence(AI).With the development of big data technology,DL ushered in another period of rapid development,which also makes the researches of depth learning coruscate new vitality.Convolutional Neural Network(CNN)is a hierarchical model to simulate the visual system which based on artificial neural network to build a hierarchical structure of the artificial neural network model.It has a great advantage in processing image recognition problems,widely used in the field of modern pattern recognition.This paper sorts out and summarizes the basic theory of deep learning at home and abroad as well as its achievements in engineering application.After analyzing the structure of CNN,an image recognition and sentence recommender system was developed based on Word2Vecand TensorFlow in the purpose of study DL in engineering fields.In this paper,the following works is done:Sort out the research results at home and abroad about DL,and analyze the background and application of deep learning;Analyze the structure and basic principle of CNN and Word2Vec,also introduce the basic algorithm to understand the network;Design the recommendation system as well as the CNN model in image recognition which based on typical CNN structures;Prepare training data set,construct DL framework,train the CNN model and development the recommender system.Based on above works,this paper verifies the advantages of deep learning in the fields of Image Recognition and Natural Language Processing from the view of project application.