Scalable high-resolution forecasting of sparse spatiotemporal events with kernel methods: a winning solution to the NIJ ``Real-time crime forecasting challenge

DOI10.1214/19-AOAS1284zbMath1435.62431arXiv1801.02858OpenAlexW2991308623MaRDI QIDQ2291539

Michael Chirico, Pau Pereira, Charles E. Loeffler, Seth R. Flaxman

Publication date: 31 January 2020

Published in: The Annals of Applied Statistics (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1801.02858

zbMATH Keywords

spatial statisticstime seriessupervised learningCox processRKHSspatiotemporal forecasting

Mathematics Subject Classification ID

Directional data; spatial statistics (62H11) Inference from stochastic processes and prediction (62M20) Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Applications of statistics to social sciences (62P25)

Related Items (6)

A Metric Space for Point Process Excitations ⋮ Hawkes Processes Modeling, Inference, and Control: An Overview ⋮ Comparative evaluation of point process forecasts ⋮ Approximate filtering of conditional intensity process for Poisson count data: application to urban crime ⋮ Scalable high-resolution forecasting of sparse spatiotemporal events with kernel methods: a winning solution to the NIJ ``Real-time crime forecasting challenge ⋮ Mixture of multivariate Gaussian processes for classification of irregularly sampled satellite image time-series

Uses Software

• glmnet
• Spearmint
• CrimeStat

Cites Work

• Unnamed Item
• Unnamed Item
• Modeling and estimation of multi-source clustering in crime and security data
• A case study competition among methods for analyzing large spatial data
• Scalable high-resolution forecasting of sparse spatiotemporal events with kernel methods: a winning solution to the NIJ ``Real-time crime forecasting challenge
• Space and circular time log Gaussian Cox processes with application to crime event data
• Spatiotemporal Prediction for Log-Gaussian Cox Processes
• Studies in the history of probability and statistics XLIX On the Matérn correlation family
• Log Gaussian Cox Processes
• Bayesian Estimation and Prediction for Inhomogeneous Spatiotemporal Log-Gaussian Cox Processes Using Low-Rank Models, With Application to Criminal Surveillance
• Rotational grid, PAI‐maximizing crime forecasts
• Autoregressive Mixture Models for Dynamic Spatial Poisson Processes: Application to Tracking Intensity of Violent Crime
• Self-Exciting Point Process Modeling of Crime
• Regularization and Variable Selection Via the Elastic Net
• Probabilistic numerics and uncertainty in computations
• Universality, Characteristic Kernels and RKHS Embedding of Measures
• Perfect simulation of Hawkes processes
• Spatial and spatio-temporal log-Gaussian Cox processes: extending the geostatistical paradigm

This page was built for publication: Scalable high-resolution forecasting of sparse spatiotemporal events with kernel methods: a winning solution to the NIJ ``Real-time crime forecasting challenge