平行学习——机器学习的一个新型理论框架

李力; 林懿伦; 曹东璞; 郑南宁; 王飞跃

doi:10.16383/j.aas.2017.y000001

平行学习——机器学习的一个新型理论框架

doi: 10.16383/j.aas.2017.y000001

李力^1,,
林懿伦^2,3,,
曹东璞^4,5,,
郑南宁^6,,
王飞跃^2,7, ,

1.
清华信息科学与技术国家实验室(筹), 清华大学自动化系北京 100084 中国
2.
中国科学院自动化研究所复杂系统管理与控制国家重点实验室北京 100190 中国
3.
中国科学院大学北京 100049 中国
4.
英国克兰菲尔德大学驾驶员认知与自动驾驶实验室克兰菲尔德 MK43 0AL 英国
5.
青岛智能产业技术研究院青岛 266000 中国
6.
西安交通大学人工智能与机器人研究所西安 710049 中国
7.
国防科技大学军事计算实验与平行系统技术中心长沙 410073 中国

基金项目:

国家自然科学基金 91520301

详细信息

作者简介:
李力清华大学自动化系副教授.主要研究方向为人工智能和机器学习,智能交通系统和智能汽车.E-mail:li-li@tsinghua.edu.cn

林懿伦中国科学院自动化研究所复杂系统管理与控制国家重点实验室博士研究生.主要研究方向为社会计算,智能交通系统,深度学习和强化学习.E-mail:linyilun2014@ia.ac.cn

曹东璞英国克兰菲尔德大学驾驶员认知与自动驾驶实验室主任.中科院自动化所客座研究员.主要研究方向为自动驾驶,人车协同,与平行驾驶.E-mail:d.cao@cranfield.ac.uk

郑南宁西安交通大学人工智能与机器人研究所教授.中国工程院院士.主要研究方向为模式识别与智能系统,机器视觉与图象处理.E-mail:nnzheng@mail.xjtu.edu.cn

通讯作者:
王飞跃中国科学院自动化研究所复杂系统管理与控制国家重点实验室研究员.国防科学技术大学军事计算实验与平行系统技术研究中心主任.主要研究方向为智能系统和复杂系统的建模、分析与控制.本文通信作者.E-mail:feiyue.wang@ia.ac.cn.

计量
- 文章访问数: 9224
- HTML全文浏览量: 2288
- PDF下载量: 9669
- 被引次数: 0
出版历程
- 收稿日期: 2016-11-25
- 录用日期: 2017-01-01
- 刊出日期: 2017-01-01

Parallel Learning——A New Framework for Machine Learning

LI Li^1
,,
LIN Yi-Lun^{2,3
,},
CAO Dong-Pu^{4,5
,},
ZHENG Nan-Ning^6
,,
WANG Fei-Yue^{2,7
, ,}

1.
National Laboratory for Information Science and Technology(TNList), Department of Automation, Tsinghua University, Beijing 100084, China
2.
the State Key Laboratory of Management and Control for Complex Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China
3.
University of Chinese Academy of Sciences, Beijing 100049, China
4.
Driver Cognition and Automated Driving Laboratory, Cranfield University, Cranfield MK43 0AL, UK
5.
Qingdao Academy of Intelligent Industries, Qingdao 266000, China
6.
Institute of Artificial Intelligence and Robotics(IAIR), Xi'an Jiaotong University, Xi'an 710049, China
7.
Research Center of Military Computational Experiments and Parallel Systems, National University of Defense Technology, Changsha 410073, China

Funds:

Supported by National Natural Science Foundation of China 91520301

More Information

Author Bio:
LI Li Associate professor at Depart- ment of Automation, Tsinghua Univer-sity. His research interest covers arti- cial intelligence and machine learning, intelligent transportation systems and intelligent vehicles.

LIN Yi-Lun Ph.D at the State Key Laboratory of Management and Con-trol for Complex Systems, Institute of Automation, Chinese Academy of Sciences. His research interest covers social computing, intelligent transportation systems, deep learning and reinforcement learning.

CAO Dong-Pu Director of Driver Cognition and Automated Driving Lab- oratory, Cran¯eld University. Visiting professor at Institute of Automation, Chinese Academy of Sciences. His research interest covers automated driving, driver-automation collaboration and parallel driving.

ZHENG Nan-Ning Professor at Institute of Arti¯cial Intelligence and Robotics (IAIR), Xi0an Jiaotong Uni- versity. Member of Chinese Academy of Engineering. His research interest covers pattern recognition and intelligent systems, computer vision and image processing.

Corresponding author: WANG Fei-Yue Professor at the State Key Laboratory of Management and Control for Complex Systems, Institute of Automation, Chinese Academy of Sciences. Director of the Research Center for Computational Experiments and Parallel Systems Technol- ogy, National University of Defense Technology. His re- search interest covers modeling, analysis, and control of intelligent systems and complex systems. Corresponding author of this paper.E-mail:feiyue.wang@ia.ac.cn.

摘要

摘要: 本文提出了一种新的机器学习理论框架.该框架结合了现有多种机器学习理论框架的优点，并针对如何使用软件定义的人工系统从大数据提取有效数据，如何结合预测学习和集成学习，以及如何利用默顿定律进行指示学习等目前机器学习领域面临的重要问题进行了特别设计.
- 机器学习 /
- 人工智能 /
- 平行学习 /
- 平行智能 /
- 平行系统及理论
Abstract: In this paper, we propose a new framework of machine learning theory, parallel learning,which incorporates and inherits many elements from various existing machine learning theories. Special designs are also presented to deal with some important problems in the machine learning research field, e.g., useful data retrieval from big data using software defined artificial systems, combination of predictive learning and ensemble learning, application of Merton's law to prescriptive learning.
- Machine learning /
- artificial intelligence /
- parallel learning /
- parallel intelligence /
- parallel system and theory

HTML全文

图 1 平行学习的理论框架图(虚线上方为通过软件定义的人工系统进行大数据预处理,虚线下方表示基于计算实验的预测学习和集成学习,以及平行控制和指示学习. 细线箭头代表数据生成或数据学习,粗线箭头代表行动和数据之间的交互.)

Fig. 1 The theoretical framework of parallel learning (The part above the dash line focuses on big data preprocessing using software defined artificial systems; the part beneath the dash line focuses on predictive learning and ensemble learning based computational experiments,as well as parallel control and prescriptive learning. The thin arrows represent either data generation or data learning; the thick arrows present interactions between data and actions.)

下载: 全尺寸图片幻灯片

图 2 AlphaGo将现实世界的数据映射到平行世界,进行多线迭代来求取预期行动

Fig. 2 AlphaGo maps data in realistic world into parallel world and uses multithread iterations to determine the expected actions

下载: 全尺寸图片幻灯片

参考文献(54)

[1]	Zheng Nan-Ning. On challenges in artificial intelligence. Acta Automatica Sinica, 2016, 42(5):641-642(郑南宁. 人工智能面临的挑战. 自动化学报, 2016, 42(5):641-642) http://cn.bing.com/academic/profile?id=7552814bad8265e3a165d5411c7423e9&encoded=0&v=paper_preview&mkt=zh-cn
[2]	王飞跃. 关于复杂系统研究的计算理论与方法. 中国基础科学, 2004, 6(5):3-10 http://www.cnki.com.cn/Article/CJFDTOTAL-ZGJB200405000.htm Wang Fei-Yue. Computational theory and method on complex system. China Basic Science, 2004, 6(5):3-10 http://www.cnki.com.cn/Article/CJFDTOTAL-ZGJB200405000.htm
[3]	王飞跃. 人工社会、计算实验、平行系统——关于复杂社会经济系统计算研究的讨论. 复杂系统与复杂性科学, 2004, 1(4):25-35 http://www.cnki.com.cn/Article/CJFDTOTAL-FZXT200404001.htm Wang Fei-Yue. Artificial societies, computational experiments, and parallel systems:a discussion on computational theory of complex social-economic systems. Complex Systems and Complexity Science, 2004, 1(4):25-35 http://www.cnki.com.cn/Article/CJFDTOTAL-FZXT200404001.htm
[4]	王飞跃. 平行系统方法与复杂系统的管理和控制. 控制与决策, 2004, 19(5):485-489 http://www.cnki.com.cn/Article/CJFDTOTAL-KZYC200405001.htm Wang Fei-Yue. Parallel system methods for management and control of complex systems. Control and Decision, 2004, 19(5):485-489 http://www.cnki.com.cn/Article/CJFDTOTAL-KZYC200405001.htm
[5]	王飞跃,史帝夫·兰森.从人工生命到人工社会——复杂社会系统研究的现状与展望.复杂系统与复杂科学, 2004, 1(1):33-41 http://www.cnki.com.cn/Article/CJFDTOTAL-FZXT200401008.htm Wang F Y, Lansing J S. From artificial life to artificial societies——new methods for studies of complex social systems. Complex Systems and Complexity Science, 2004, 1(1):33-41 http://www.cnki.com.cn/Article/CJFDTOTAL-FZXT200401008.htm
[6]	Zhang N, Wang F Y, Zhu F H, Zhao D B, Tang S M. DynaCAS:computational experiments and decision support for ITS. IEEE Intelligent Systems, 2008, 23(6):19-23 doi: 10.1109/MIS.2008.101
[7]	Wang F Y. Parallel control and management for intelligent transportation systems:concepts, architectures, and applications. IEEE Transactions on Intelligent Transportation Systems, 2010, 11(3):630-638 doi: 10.1109/TITS.2010.2060218
[8]	Li L, Wen D. Parallel systems for traffic control:a rethinking. IEEE Transactions on Intelligent Transportation Systems, 2016, 17(4):1179-1182 doi: 10.1109/TITS.2015.2494625
[9]	Wang F Y, Wang X, Li L X, Li L. Steps toward parallel intelligence. IEEE/CAA Journal of Automatica Sinica, 2016, 3(4):345-348 doi: 10.1109/JAS.2016.7510067
[10]	Jain A K, Duin R P W, Mao J C. Statistical pattern recognition:a review. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(1):4-37 doi: 10.1109/34.824819
[11]	Bishop C M. Pattern Recognition and Machine Learning. New York:Springer, 2006.
[12]	Shalev-Shwartz S. Online learning and online convex optimization. Foundations and Trends in Machine Learning, 2011, 4(2):107-194 doi: 10.1561/2200000018
[13]	Rakhlin A, Sridharan K. Statistical learning and sequential prediction[Online], available:http://www-stat.wharton.upenn.edu/rakhlin/courses/stat928/stat928_notes.pdf, January 1, 2017
[14]	Pérez-Sánchez B, Fontenla-Romero O, Guijarro-Berdiñas B. A review of adaptive online learning for artificial neural networks. Artificial Intelligence Review, 2016, DOI: 10.1007/s10462-016-9526-2
[15]	Wiering M, van Otterlo M. Reinforcement Learning:State-of-the-Art. Berlin Heidelberg:Springer, 2012.
[16]	Sutton R S, Barto A G. Reinforcement Learning:An Introduction (Second edition). Cambridge, MA:The MIT Press, 2017.
[17]	Settles B. Active Learning. San Rafael, CA:Morgan&Claypool Publishers, 2012.
[18]	Wang F Y. A big-data perspective on AI:Newton, Merton, and analytics intelligence. IEEE Intelligent Systems, 2012, 27(5):2-4 doi: 10.1109/MIS.2012.91
[19]	Zeng D, Lusch R. Big data analytics:perspective shifting from transactions to ecosystems. IEEE Intelligent Systems, 2013, 28(2):2-5 doi: 10.1109/MIS.2013.40
[20]	Zhou Z H, Chawla N V, Jin Y C, Williams G J. Big data opportunities and challenges:discussions from data analytics perspectives. IEEE Computational Intelligence Magazine, 2014, 9(4):62-74 doi: 10.1109/MCI.2014.2350953
[21]	Richard Feynman[Online], available:https://en.wikiquote.org/wiki/Richard_Feynman, January 1, 2017
[22]	Bainbridge W S. The scientific research potential of virtual worlds. Science, 2007, 317(5837):472-476 doi: 10.1126/science.1146930
[23]	王坤峰, 苟超, 王飞跃. 平行视觉:基于ACP的智能视觉计算方法. 自动化学报, 2016, 42(10):1490-1500 http://www.aas.net.cn/CN/abstract/abstract18936.shtml Wang Kun-Feng, Gou Chao, Wang Fei-Yue. Parallel vision:an ACP-based approach to intelligent vision computing. Acta Automatica Sinica, 2016, 42(10):1490-1500 http://www.aas.net.cn/CN/abstract/abstract18936.shtml
[24]	Gaidon A, Wang Q, Cabon Y, Vig E. Virtual worlds as proxy for multi-object tracking analysis. In:Proceedings of the 2016 Computer Vision and Pattern Recognition. Las Vegas, NV:CVPR, 2016. 4340-4349
[25]	Wang K, Gou C, Wang F Y, Rehg J M. Parallel vision for perception and understanding of complex scenes:methods, framework, and perspectives. Pattern Recognition, to be published
[26]	田渊栋. 阿法狗围棋系统的简要分析. 自动化学报, 2016, 42(5):671-675 http://www.aas.net.cn/CN/abstract/abstract18856.shtml Tian Yuan-Dong. A simple analysis of AlphaGo. Acta Automatica Sinica, 2016, 42(5):671-675 http://www.aas.net.cn/CN/abstract/abstract18856.shtml
[27]	Wang F Y, Zhang J J, Zheng X H, Wang X, Yuan Y, Dai X X, Zhang J, Yang L Q. Where does AlphaGo go:from church-turing thesis to AlphaGo thesis and beyond. IEEE/CAA Journal of Automatica Sinica, 2016, 3(2):113-120 doi: 10.1109/JAS.2016.7471613
[28]	Silver D, Huang A, Maddison C J, Guez A, Sifre L, van den Driessche G, Schrittwieser J, Antonoglou I, Panneershelvam V, Lanctot M, Dieleman S, Grewe D, Nham J, Kalchbrenner N, Sutskever I, Lillicrap T, Leach M, Kavukcuoglu K, Graepel T, Hassabis D. Mastering the game of Go with deep neural networks and tree search. Nature, 2016, 529(7587):484-489 doi: 10.1038/nature16961
[29]	Hinton G, Sejnowski T J. Unsupervised Learning:Foundations of Neural Computation. Cambridge:MIT Press, 1999.
[30]	Duda R O, Hart P E, Stork D G. Unsupervised learning and clustering. Pattern Classification (Second edition). New York:Wiley, 2001. 517-600
[31]	LeCun Y. Predictive Learning. Speech at NIPS, 2016. https://drive.google.com/file/d/0BxKBnD5y2M8NREZod0tVdW5FLTQ/view
[32]	Piaget J. The Origins of Intelligence in Children (Second edition). New York:International Universities Press Inc., 1974.
[33]	Drescher G L. Made-Up Minds:A Constructivist Approach to Artificial Intelligence. Cambridge:MIT Press, 1991.
[34]	Hawkins J, Blakeslee S. On Intelligence:How a New Understanding of the Brain will Lead to the Creation of Truly Intelligent Machines. New York:Times Books, 2004.
[35]	Zhu X J, Goldberg A B. Introduction to Semi-Supervised Learning. San Francisco, CA:Morgan&Claypool Publishers, 2009.
[36]	Schwenker F, Trentin E. Pattern classification and clustering:a review of partially supervised learning approaches. Pattern Recognition Letters, 2014, 37:4-14 doi: 10.1016/j.patrec.2013.10.017
[37]	Minsky M. The Society of Mind. New York:Simon&Schuster, 1988.
[38]	Shoham Y, Leyton-Brown K. Multiagent Systems:Algorithmic, Game-Theoretic, and Logical Foundations. Cambridge:Cambridge University Press, 2008.
[39]	Zhou Z H. Ensemble Methods:Foundations and Algorithms. Boca Raton, FL, USA:CRC Press, 2012.
[40]	Li L, Chen X Q, Zhang L. Multimodel ensemble for freeway traffic state estimations. IEEE Transactions on Intelligent Transportation Systems, 2014, 15(3):1323-1336 doi: 10.1109/TITS.2014.2299542
[41]	Wang F Y, Zhang J, Wei Q, Li L. PDP:parallel dynamic programming. IEEE/CAA Journal of Automatica Sinica, 2017, 4(1):1-5 doi: 10.1109/JAS.2017.7510310
[42]	Merton R K. The unanticipated consequences of purposive social action. American Sociological Review, 1936, 1(6):894-904 doi: 10.2307/2084615
[43]	王飞跃. 软件定义的系统与知识自动化:从牛顿到默顿的平行升华. 自动化学报, 2015, 41(1):1-8 http://www.aas.net.cn/CN/abstract/abstract18578.shtml Wang Fei-Yue. Software-defined systems and knowledge automation:a parallel paradigm shift from Newton to Merton. Acta Automatica Sinica, 2015, 41(1):1-8 http://www.aas.net.cn/CN/abstract/abstract18578.shtml
[44]	Lowd D, Meek C. Adversarial learning. In:Proceedings of the 7th ACM SIGKDD International Conference on Knowledge Discovery in Data Mining. Chicago, Illinois:ACM, 2005.
[45]	Goodfellow I J, Pouget-Abadie J, Mirza M, Xu B, Warde-Farley D, Ozair S, Courville A, Bengio Y. Generative adversarial nets. In:Proceedings of Advances in Neural Information Processing Systems 27:28th Annual Conference on Neural Information Processing Systems 2014. Montreal, Canada:NIPS, 2014. 2672-2680
[46]	Goodfellow I J, Shlens J, Szegedy C. Explaining and harnessing adversarial examples[Online], available:https://arxiv.org/abs/1412.6572, January 1, 2017
[47]	Xia Y C, He D, Qin T, Wang L W, Yu N H, Liu T Y, Ma W Y. Dual learning for machine translation[Online], available:https://arxiv.org/abs/1611.00179, January 1, 2017
[48]	Ullrich C. Descriptive and prescriptive learning theories. Pedagogically Founded Courseware Generation for Web-Based Learning:Lecture Notes in Computer Science. Berlin Heidelberg:Springer, 2008, 5260:37-42
[49]	Chen X, Duan Y, Houthooft R, Schulman J, Sutskever I, Abbeel P. InfoGAN:Interpretable representation learning by information maximizing generative adversarial nets[Online], available:https://arxiv.org/abs/1606.03657, January 1, 2017
[50]	Li Y C, Liu Y H, Zhang C, Zhao D C, Zheng N N. The "floor-wall" traffic scenes construction for unmanned vehicle simulation evaluation. In:Proceedings of the 17th IEEE Conference on Intelligent Transportation Systems. Qingdao, China:IEEE, 2014. 1726-1731
[51]	Li L, Huang W L, Liu Y H, Zheng N N, Wang F Y. Intelligence testing for autonomous vehicles:a new approach. IEEE Transactions on Intelligent Vehicles, 2016, 1(2):158-166 doi: 10.1109/TIV.2016.2608003
[52]	王飞跃. 社会信号处理与分析的基本框架:从社会传感网络到计算辩证解析方法. 中国科学:信息科学, 2013, 43(12):1598-1611 http://www.cnki.com.cn/Article/CJFDTOTAL-PZKX201312008.htm Wang Fei-Yue. A framework for social signal processing and analysis:from social sensing networks to computational dialectical analytics. Scientia Sinica Informationis, 2013, 43(12):1598-1611 http://www.cnki.com.cn/Article/CJFDTOTAL-PZKX201312008.htm
[53]	王飞跃. 平行控制:数据驱动的计算控制方法. 自动化学报, 2013, 39(4):293-302 http://www.aas.net.cn/CN/abstract/abstract17915.shtml Wang Fei-Yue. Parallel control:a method for data-driven and computational control. Acta Automatica Sinica, 2013, 39(4):293-302 http://www.aas.net.cn/CN/abstract/abstract17915.shtml
[54]	王飞跃. 天命唯新:迈向知识自动化——《自动化学报》创刊50周年专刊序. 自动化学报, 2013, 39(11):1741-1743 doi: 10.3724/SP.J.1004.2013.01741 Wang Fei-Yue. The destiny:towards knowledge automation——preface of the special issue for the 50th anniversary of Acta Automatica Sinica. Acta Automatica Sinica, 2013, 39(11):1741-1743 doi: 10.3724/SP.J.1004.2013.01741