【IJAC热文】MIT Tomaso Poggio教授探讨深度学习机理
上周，在IJAC优先在线发表的几篇论文中，麻省理工美国人文与科学院院士Tomaso Poggio的一篇有关深度学习的综述成为一大亮点。Poggio教授在这篇文章中阐述了有关神经网络深度学习的一个基本问题：为什么深层网络比浅层网络更好？
文章内容延续了Poggio教授在2016年8月在中国人工智能大会（2016 CCAI）上的演讲报告《The Science and the Engineering of Intelligence》。
图1 来源2016 CCAI 演讲PPT
图2 来源2016 CCAI 演讲PPT
深度学习架构和机器学习模式的搭建，来自于神经学方面的研究进展，换句话说，同样的架构是存在于大脑皮质当中的。
关于深度学习，已经有成千上万的研究者在不同领域进行这方面的研究，比如无人驾驶、语音识别等等。可是我们还不清楚，为什么深度学习在这些工程应用中会起作用，深度学习的机理是什么？
我们对这个问题很感兴趣另外一个原因是：探讨深度学习的机理也将有助于我们理解‘为什么大脑皮质会存在一些不同的层次’。
”Poggio教授在这篇文章中，将为您解读深度学习的关键理论、最新成果和开放式研究问题。
同时这篇文章也是IJAC即将发表的Special Issue on Human Inspired Computing中的一篇文章。该专题其他热文将陆续优先在线发表，敬请期待。
一点点题外话：小编在去年的CCAI大会上有幸拜访了Poggio教授，教授博学、谦逊而富有亲和力的形象给小编也留下深刻印象。他曾提到：期望能帮助年轻人更好的了解神经科学、理解机器学习。如果要在智能方面走得远，不能只靠计算机，还需要与人类本身的研究相互结合，才能碰撞出更多的东西。
接下来，小编为您奉上文章的详细信息，
欢迎下载阅读：
【Title】Why and when can deep-but not shallow-networks avoid the curse of dimensionality: A review
【Authors】Tomaso Poggio, Hrushikesh Mhaskar, Lorenzo Rosasco, Brando Miranda, Qianli Liao
【Abstract】The paper reviews and extends an emerging body of theoretical results on deep learning including the conditions under which it can be exponentially better than shallow learning. A class of deep convolutional networks represent an important special case of these conditions, though weight sharing is not the main reason for their exponential advantage. Implications of a few key theorems are discussed, together with new results, open problems and conjectures.
【Keywords】Machine learning, neural networks, deep and shallow networks, convolutional neural networks, function approximation, deep learning
【Full Text】
3月上旬，IJAC优先在线发表的论文还有如下，欢迎阅读
【Title】Reactive navigation of underwater mobile robot using ANFIS approach in a manifold manner
【Authors】Shubhasri Kundu, Dayal R. Parhi
【Keywords】Adaptive fuzzy inference system (ANFIS), error gradient, optimal path, obstacle avoidance behavior, steering angle, target seeking behavior
【Full Text】
https://link.springer.com/article/10.1007/s11633-016-0983-5
【Title】Reaction torque control of redundant free-floating space robot
【Authors】Ming-He Jin, Cheng Zhou, Ye-Chao Liu, Zi-Qi Liu, Hong Liu
【Keywords】Redundant space robot, reaction torque, reactionless control, base disturbance minimization, Linux/real time application interface (RTAI)
【Full Text】
https://link.springer.com/article/10.1007/s11633-017-1067-x
【Title】Navigation of non-holonomic mobile robot using neuro-fuzzy logic with integrated safe boundary algorithm
【Authors】A. Mallikarjuna Rao, K. Ramji, B. S. K. Sundara Siva Rao, V. Vasu, C. Puneeth
【Keywords】Robotics, autonomous mobile robot (AMR), navigation, fuzzy logic, neural networks, adaptive neuro-fuzzy inference system (ANFIS), safe boundary algorithm
【Full Text】
https://link.springer.com/article/10.1007/s11633-016-1042-y
【Title】Virtual plate based controlling strategy of toy play for robots communication development in JA space
【Authors】Wei Wang, Xiao-Dan Huang
【Keywords】Human robot cooperation, joint attention (JA) space, reachable space, toy play ability, a virtual plate
【Full Text】
https://link.springer.com/article/10.1007/s11633-016-1022-2
【Title】Method for visual localization of oil and gas wellhead based on distance function of projected features
【Authors】Ying Xie, Xiang-Dong Yang, Zhi Liu, Shu-Nan Ren, Ken Chen
【Keywords】Robot vision, visual localization, 3D object localization, model based pose estimation, distance function of projected features, nonlinear least squares, random sample consensus (RANSAC)
【Full Text】
https://link.springer.com/article/10.1007/s11633-017-1063-1
【Title】A piecewise switched linear approach for traffic flow modeling
【Authors】Abdelhafid Zeroual,Nadhir Messai, Sihem Kechida, Fatiha Hamdi
【Keywords】Switched systems, modeling, macroscopic, traffic flow, data calibration
【Full Text】
https://link.springer.com/article/10.1007/s11633-017-1060-4
【Title】Improvement of wired drill pipe data quality via data validation and reconciliatio
【Authors】Dan Sui, Olha Sukhoboka, Bernt Sigve Aadnøy
【Keywords】Data quality, wired drill pipe (WDP), data validation and reconciliation (DVR), drilling models
【Full Text】
https://link.springer.com/article/10.1007/s11633-017-1068-9
阅读更多IJAC OnlineFirst Papers：
点击阅读相关文章
更多精彩内容，敬请关注：
IJAC官方网站：
1) https://link.springer.com/journal/11633
2)
微信公众号：IJAC
新浪微博：IJAC-国际自动化与计算杂志
Twitter: IJAC_Journal
Linked in (领英):
Int. J. of Automation and Computing
微信公众号：IJAC
International Journal of
Automation and Computing
IJAC官方微信平台为您呈现期刊在线、最新资讯等栏目，致力于促进学术作品的出版、分享与传播。