Data-driven chimney fire risk prediction using machine learning and point process tools
From MaRDI portal
Publication:6138624
DOI10.1214/23-AOAS1752arXiv2112.07257OpenAlexW4388076507MaRDI QIDQ6138624FDOQ6138624
Authors: Changqing Lü, M. N. M. Van Lieshout, Maurits de Graaf, Paul Visscher 
Publication date: 16 January 2024
Published in: The Annals of Applied Statistics (Search for Journal in Brave)
Abstract: Chimney fires constitute one of the most commonly occurring fire types. Precise prediction and prompt prevention are crucial in reducing the harm they cause. In this paper, we develop a combined machine learning and statistical modeling process to predict chimney fires. Firstly, we use random forests and permutation importance techniques to identify the most informative explanatory variables. Secondly, we design a Poisson point process model and apply associated logistic regression estimation to estimate the parameters. Moreover, we validate the Poisson model assumption using second-order summary statistics and residuals. We implement the modeling process on data collected by the Twente Fire Brigade and obtain plausible predictions. Compared to similar studies, our approach has two advantages: i) with random forests, we can select explanatory variables non-parametrically considering variable dependence; ii) using logistic regression estimation, we can fit the statistical model efficiently by tuning it to focus on important regions and times of the fire data.
Full work available at URL: https://arxiv.org/abs/2112.07257
variable importancepair correlation functionPoisson point process\(K\)-functionfire predictionlogistic regression estimationspatiotemporal point pattern
Cites Work
- Statistical Analysis of Financial Data in S-Plus
- Random forests
- Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations (with discussion)
- On the Second-Order and Orientation Analysis of Planar Stationary Point Processes
- Consistency and asymptotic normality of the maximum likelihood estimator in generalized linear models
- Title not available (Why is that?)
- Survival ensembles
- Quasi-Likelihood and Optimal Estimation, Correspondent Paper
- Who Invented the Delta Method?
- An Introduction to the Theory of Point Processes
- Title not available (Why is that?)
- Practical Maximum Pseudolikelihood for Spatial Point Patterns
- Residual Analysis for Spatial Point Processes (with Discussion)
- Structured spatio-temporal shot-noise Cox point process models, with a view to modeling forest fires
- Point process modeling of wildfire hazard in Los Angeles county, California
- Variable selection for spatial Poisson point processes via a regularization method
- Logistic regression for spatial Gibbs point processes
- Variable selection for inhomogeneous spatial point process models
- Convex and non-convex regularization methods for spatial point processes intensity estimation
- Quantification of annual wildfire risk; A spatio-temporal point process approach.
- Spatiotemporal wildfire modeling through point processes with moderate and extreme marks
- Theory of Spatial Statistics
- Information criteria for inhomogeneous spatial point processes
- Enhancing the SPDE modeling of spatial point processes with INLA, applied to wildfires. Choosing the best mesh for each database
This page was built for publication: Data-driven chimney fire risk prediction using machine learning and point process tools
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6138624)
