PAC Learnability of Node Functions in Networked Dynamical Systems

We consider the PAC learnability of the functions at the nodes of a discrete networked dynamical system, assuming that the underlying network is known. We provide tight bounds on the sample complexity of learning threshold functions. We establish a computational intractability result for efficient PAC learning of such functions. We develop efficient consistent learners when the number of negative examples is small. Using synthetic and real-world networks, we experimentally study how the network structure and sample complexity influence the quality of inference.

[1]  Thomas G. Dietterich.  What is machine learning? , 2020, Archives of Disease in Childhood.

[2]  Tanya Y. Berger-Wolf,et al.  Network Structure Inference, A Survey , 2016, ACM Comput. Surv..

[3]  T. Valente.  Social network thresholds in the diffusion of innovations , 1996 .

[4]  Yamir Moreno,et al.  The Dynamics of Protest Recruitment through an Online Network , 2011, Scientific reports.

[5]  D. Angluin,et al.  Randomly fallible teachers: Learning monotone DNF with an incomplete membership oracle , 1991, Machine Learning.

[6]  Mark S. Granovetter.  Threshold Models of Collective Behavior , 1978, American Journal of Sociology.

[7]  Madhav V. Marathe,et al.  Learning the Behavior of a Dynamical System Via a "20 Questions" Approach , 2018, AAAI.

[8]  Samarth Swarup,et al.  Inhibiting the Diffusion of Contagions in Bi-Threshold Systems: Analytical and Experimental Results , 2011, AAAI Fall Symposium: Complex Adaptive Systems.

[9]  David Haussler,et al.  Quantifying Inductive Bias: AI Learning Algorithms and Valiant's Learning Framework , 1988, Artif. Intell..

[10]  Duncan J Watts,et al.  A simple model of global cascades on random networks , 2002, Proceedings of the National Academy of Sciences of the United States of America.

[11]  Christian M. Reidys,et al.  An Introduction to Sequential Dynamical Systems , 2007, Universitext.

[12]  Rocco A. Servedio,et al.  On PAC learning algorithms for rich Boolean function classes , 2006, Theoretical Computer Science.

[13]  Philip M. Long.  On the sample complexity of PAC learning half-spaces against the uniform distribution , 1995, IEEE Trans. Neural Networks.

[14]  T. Schelling.  Micromotives and Macrobehavior , 1978 .

[15]  Joel C. Miller.  Spread of infectious disease through clustered populations , 2008, Journal of The Royal Society Interface.

[16]  Éva Tardos,et al.  Maximizing the Spread of Influence through a Social Network , 2015, Theory Comput..

[17]  David Haussler,et al.  Learnability and the Vapnik-Chervonenkis dimension , 1989, JACM.

[18]  Abhijin Adiga,et al.  Inferring local transition functions of discrete dynamical systems from observations of system behavior , 2017, Theor. Comput. Sci..

[19]  Umesh V. Vazirani,et al.  An Introduction to Computational Learning Theory , 1994 .

[20]  Jan Vondrák,et al.  Maximizing a Monotone Submodular Function Subject to a Matroid Constraint , 2011, SIAM J. Comput..

[21]  Martin Hoefer,et al.  Learning Market Parameters Using Aggregate Demand Queries , 2016, AAAI.

[22]  Andrey Y. Lokhov,et al.  Reconstructing Parameters of Spreading Models from Partial Observations , 2016, NIPS.

[23]  Leslie G. Valiant,et al.  A theory of the learnable , 1984, CACM.

[24]  Manfred K. Warmuth.  Towards Representation Independence in PAC Learning , 1989, AII.

[25]  Yan Liu,et al.  Learning Influence Functions from Incomplete Observations , 2016, NIPS.

[26]  M. Newman.  Spread of epidemic disease on networks. , 2002, Physical review. E, Statistical, nonlinear, and soft matter physics.

[27]  Nikolaos Fountoulakis,et al.  Bootstrap Percolation in Power-Law Random Graphs , 2011, Journal of Statistical Physics.

[28]  Jon Kleinberg,et al.  Differences in the mechanics of information diffusion across topics: idioms, political hashtags, and complex contagion on twitter , 2011, WWW.

[29]  David C. Parkes,et al.  Learnability of Influence in Networks , 2015, NIPS.

[30]  Steve Hanneke,et al.  The Optimal Sample Complexity of PAC Learning , 2015, J. Mach. Learn. Res..

[31]  Cécile Favre,et al.  Information diffusion in online social networks: a survey , 2013, SGMD.