本文選題：深度學(xué)習(xí)　切入點(diǎn)：卷積神經(jīng)網(wǎng)絡(luò)　出處：《北京交通大學(xué)》2017年碩士論文　論文類型：學(xué)位論文

【摘要】：深度學(xué)習(xí)(DL,Deep Learning)是機(jī)器學(xué)習(xí)(ML,Machine Learning)的一個(gè)重要方法和研究方向,屬于人工智能(AI,Artificial Intelligence)領(lǐng)域的重要分支。隨著大數(shù)據(jù)技術(shù)的發(fā)展,深度學(xué)習(xí)迎來(lái)了又一個(gè)快速發(fā)展的時(shí)期,這也使得深度學(xué)習(xí)理論與算法研究煥發(fā)新的活力。卷積神經(jīng)網(wǎng)絡(luò)(CNN,Convolutional Neural Network)作為深度學(xué)習(xí)模型的代表,是模擬視覺(jué)系統(tǒng)層次化的工作模式,在人工神經(jīng)網(wǎng)絡(luò)的基礎(chǔ)上構(gòu)建具有層次化結(jié)構(gòu)的人工網(wǎng)絡(luò)模型。其局部感知、層次結(jié)構(gòu)化等特點(diǎn)在處理圖像識(shí)別問(wèn)題上具有巨大優(yōu)勢(shì),在現(xiàn)代模式識(shí)別領(lǐng)域獲得了廣泛應(yīng)用。本文在整理與總結(jié)國(guó)內(nèi)外深度學(xué)習(xí)的基本理論成果與在工程上的應(yīng)用現(xiàn)狀,并對(duì)卷積神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu)分析的基礎(chǔ)上,結(jié)合Word2Vec與TensorFlow深度學(xué)習(xí)框架,開(kāi)發(fā)了圖像識(shí)別與文字推薦系統(tǒng),以工程應(yīng)用為背景對(duì)其理論成果進(jìn)行研究。本文主要進(jìn)行了以下幾項(xiàng)工作:整理國(guó)內(nèi)外深度學(xué)習(xí)的研究成果,并對(duì)深度學(xué)習(xí)的背景與應(yīng)用進(jìn)行總結(jié);分析卷積神經(jīng)網(wǎng)絡(luò)與Word2Vec的結(jié)構(gòu)與基本原理,并對(duì)理解網(wǎng)絡(luò)模型所需的基本算法進(jìn)行了介紹;設(shè)計(jì)本文的圖像識(shí)別與文字推薦系統(tǒng),并以經(jīng)典CNN網(wǎng)絡(luò)結(jié)構(gòu)為基礎(chǔ)設(shè)計(jì)基于本文推薦的卷積神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu);進(jìn)行數(shù)據(jù)集的準(zhǔn)備、深度學(xué)習(xí)框架的搭建及本文模型訓(xùn)練工作,并實(shí)現(xiàn)本文圖像識(shí)別與文字推薦系統(tǒng);通過(guò)以上工作,本文從工程項(xiàng)目應(yīng)用的角度驗(yàn)證了深度學(xué)習(xí)在圖像識(shí)別與自然語(yǔ)言處理問(wèn)題上的優(yōu)勢(shì)。
[Abstract]:Deep learning is an important method and research direction of machine learning, which belongs to the important branch of artificial intelligence. With the development of big data technology, deep learning has ushered in another period of rapid development. As a representative of depth learning model, convolution neural network (CNN) is a hierarchical working mode of simulating visual system. An artificial network model with hierarchical structure is constructed on the basis of artificial neural network. Its local perception and hierarchical structure have great advantages in image recognition. It has been widely used in the field of modern pattern recognition. In this paper, the basic theoretical achievements of deep learning at home and abroad and the present situation of application in engineering are summarized, and the network structure of convolutional neural network is analyzed. Based on the Word2Vec and TensorFlow deep learning framework, an image recognition and character recommendation system is developed, and its theoretical results are studied under the background of engineering application. The main work of this paper is as follows: sorting out the research results of deep learning at home and abroad. The background and application of deep learning are summarized, the structure and basic principle of convolutional neural network and Word2Vec are analyzed, and the basic algorithms for understanding the network model are introduced. Based on the classical CNN network structure, we design the convolutional neural network structure recommended in this paper, prepare the data set, build the deep learning framework and train the model in this paper, and realize the image recognition and text recommendation system in this paper. Through the above work, this paper verifies the advantages of deep learning in image recognition and natural language processing from the point of view of engineering project application.
【學(xué)位授予單位】：北京交通大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TP391.41;TP391.3

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 楊志義;朱婭婷;蒲勇;;基于統(tǒng)一計(jì)算設(shè)備架構(gòu)技術(shù)的并行圖像處理研究[J];計(jì)算機(jī)測(cè)量與控制;2009年04期

2 沈艷軍,汪秉文,胡曉婭;多層前向神經(jīng)網(wǎng)絡(luò)帶正則化因子的算法[J];系統(tǒng)工程與電子技術(shù);2004年09期

，

本文編號(hào)：1581352