摘要
深度学习是机器学习中接近AI的领域之一,通过模拟人脑学习神经进行分析。深度学习源于人工神经网络的研究,其对比简单学习来讲,多数分类、回归等学习算法归于简单机器学习,复杂函数运算的表示能力和局限性与有限样本和计算单元对有关,泛化能力也受到复杂分类的一定限制。
Deep learning is one of the fields very close to AI in machine learning, which is by simulating the human brain to analyze learning nerve. Deep learning stems from the study of artificial neural networks, compared with simple learning, most classification, regression learning algorithms are attributed to simple machine learning, the representation and limitations of complex function operations are related to finite samples and computational units, generalization ability is also limited by complex classification.
作者
竺宝宝
张娜
Zhu Baobao Zhang Na(Henan University of Urban Construction. Pingdingshan 467036, Chin)
出处
《无线互联科技》
2017年第10期25-26,共2页
Wireless Internet Technology
关键词
深度学习
自然语言
非线性网络结构
deep learning: natural language: nonlinear network structure