No-regret Algorithms for Capturing Events in Poisson Point Processes

Inhomogeneous Poisson point processes are widely used models of event occurrences. We address adaptive sensing of Poisson Point processes, namely, maximizing the number of captured events subject to sensing costs. We encode prior assumptions on the rate function by modeling it as a member of a known reproducing kernel Hilbert space (RKHS). By partitioning the domain into separate small regions, and using heteroscedastic linear regression, we propose a tractable estimator of Poisson process rates for two feedback models: count-record, where exact locations of events are observed, and histogram feedback, where only counts of events are observed. We derive provably accurate anytime confidence estimates for our estimators for sequentially acquired Poisson count data. Using these, we formulate algorithms based on optimism that provably incur sublinear count-regret. We demonstrate the practicality of the method on problems from crime modeling, revenue maximization as well as environmental monitoring.

[1]  H. Robbins,et al.  A Class of Stopping Rules for Testing Parametric Hypotheses , 1985 .

[2]  T. Lai,et al.  Theory and applications of multivariate self-normalized processes , 2009 .

[3]  Andres Muñoz Medina,et al.  No-Regret Algorithms for Heavy-Tailed Linear Bandits , 2016, ICML.

[4]  Hassan Maatouk,et al.  Gaussian Process Emulators for Computer Experiments with Inequality Constraints , 2016, Mathematical Geosciences.

[5]  N. Cressie,et al.  Fixed rank kriging for very large spatial data sets , 2008 .

[6]  Dávid Papp,et al.  Shape-Constrained Estimation Using Nonnegative Splines , 2014 .

[7]  Elad Hazan,et al.  Logarithmic regret algorithms for online convex optimization , 2006, Machine Learning.

[8]  Andreas Krause,et al.  Information Directed Sampling and Bandits with Heteroscedastic Noise , 2018, COLT.

[9]  Marc Abeille,et al.  Improved Optimistic Algorithms for Logistic Bandits , 2020, ICML.

[10]  Stephen J. Roberts,et al.  Variational Inference for Gaussian Process Modulated Poisson Processes , 2014, ICML.

[11]  Andreas Krause,et al.  Efficient High Dimensional Bayesian Optimization with Additivity and Quadrature Fourier Features , 2018, NeurIPS.

[12]  Andreas Krause,et al.  Discovering Valuable items from Massive Data , 2015, KDD.

[13]  Andrés F. López-Lopera,et al.  Gaussian Process Modulated Cox Processes under Linear Inequality Constraints , 2019, AISTATS.

[14]  Andreas Krause,et al.  Information-Theoretic Regret Bounds for Gaussian Process Optimization in the Bandit Setting , 2009, IEEE Transactions on Information Theory.

[15]  Tor Lattimore,et al.  Information Directed Sampling for Linear Partial Monitoring , 2020, COLT.

[16]  Csaba Szepesvari,et al.  Bandit Algorithms , 2020 .

[17]  Francis R. Bach,et al.  On the Equivalence between Kernel Quadrature Rules and Random Feature Expansions , 2015, J. Mach. Learn. Res..

[18]  Gábor Rudolf,et al.  Arrival rate approximation by nonnegative cubic splines , 2005, 2005 IEEE International Conference on Electro Information Technology.

[19]  P. Diggle,et al.  Estimating weighted integrals of the second-order intensity of a spatial point process , 1989 .

[20]  Alan E. Gelfand,et al.  Space and circular time log Gaussian Cox processes with application to crime event data , 2016, 1611.08719.

[21]  Alessandro Rudi,et al.  Non-parametric Models for Non-negative Functions , 2020, NeurIPS.

[22]  Peter J. Diggle,et al.  Spatial and spatio-temporal Log-Gaussian Cox processes:extending the geostatistical paradigm , 2013, 1312.6536.

[23]  Virgilio Gómez-Rubio,et al.  Spatial Point Patterns: Methodology and Applications with R , 2016 .

[24]  Jon D. McAuliffe,et al.  Time-uniform Chernoff bounds via nonnegative supermartingales , 2018, Probability Surveys.

[25]  J. Heikkinen,et al.  Modeling a Poisson forest in variable elevations: a nonparametric Bayesian approach. , 1999, Biometrics.

[26]  Donald L. Snyder,et al.  Random Point Processes in Time and Space , 1991 .

[27]  Alexis Boukouvalas,et al.  Adaptive Sensor Placement for Continuous Spaces , 2019, ICML.

[28]  Adam N. Letchford,et al.  Adaptive Policies for Perimeter Surveillance Problems , 2018, Eur. J. Oper. Res..

[29]  Csaba Szepesvári,et al.  Improved Algorithms for Linear Stochastic Bandits , 2011, NIPS.

[30]  Peter Auer,et al.  Using Confidence Bounds for Exploitation-Exploration Trade-offs , 2003, J. Mach. Learn. Res..

[31]  Aurélien Garivier,et al.  Parametric Bandits: The Generalized Linear Case , 2010, NIPS.

[32]  Eric R. Ziegel,et al.  Generalized Linear Models , 2002, Technometrics.

[33]  Tamás Linder,et al.  The On-Line Shortest Path Problem Under Partial Monitoring , 2007, J. Mach. Learn. Res..

[34]  J. T. Wulu,et al.  Regression analysis of count data , 2002 .

[35]  Michael R. Lyu,et al.  Almost Optimal Algorithms for Linear Stochastic Bandits with Heavy-Tailed Payoffs , 2018, NeurIPS.

[36]  Sattar Vakili,et al.  On Information Gain and Regret Bounds in Gaussian Process Bandits , 2020, AISTATS.

[37]  Benjamin Van Roy,et al.  Learning to Optimize via Posterior Sampling , 2013, Math. Oper. Res..

[38]  Bernhard Schölkopf,et al.  A Generalized Representer Theorem , 2001, COLT/EuroCOLT.

[39]  Ryan P. Adams,et al.  Tractable nonparametric Bayesian inference in Poisson processes with Gaussian process intensities , 2009, ICML '09.

[40]  James Hensman,et al.  Large-Scale Cox Process Inference using Variational Fourier Features , 2018, ICML.

[41]  M. R. Leadbetter.  Poisson Processes , 2011, International Encyclopedia of Statistical Science.

[42]  Olivier Roustant,et al.  Finite-dimensional Gaussian approximation with linear inequality constraints , 2017, SIAM/ASA J. Uncertain. Quantification.

[43]  Jürgen Symanzik,et al.  Statistical Analysis of Spatial Point Patterns , 2005, Technometrics.