深度学习:开启大数据时代的钥匙

随着大数据时代的到来,基于深度学习技术的机器学习方法被用于有效地分析和处理这些数据。本文首先概述了深度学习技术的由来,对比了浅层结构与深度结构模型的差异,分析了深度结构模型在大数据应用中的优势;认为深度学习取得成功的条件是,大规模训练数据集的支撑、先进的硬件平台支持、新的优化技术的提出;基于计算机视觉应用,从有监督特征学习和无监督特征学习两个方面分别介绍了当前深度学习研究的现状和典型的深度结构模型的基本原理和主要应用;针对当前深度学习的发展现状,总结了深度学习研究存在的挑战和未来的研究方向。

Deep Learning:A key of Stepping into the Era of Big Data

With the big data era＆#39;s coming, deep learning, a new research field of machine learning, is used to analyze and process data efficiently. Firstly we introduce the history of deep leaning development, and compare the differences between shallow and deep structure models. Then we describe the advantages of deep structure model and summarize that the success of deep learning lies in three aspects：big scale data sets, advanced hard-ware facilities and new optimized techniques. Secondly, from the perspective of supervised and unsupervised learning setting, we review the current situation of deep learning research and the basic principles and applications of classical deep structure models Finally, we summarize the challenges and the direction of future research of deep learning.