Technology Category
Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal DataPlanning, Routing, and Scheduling: Optimization of Spatio-temporal SystemsMachine Learning: Time-Series/Data Streams
Autocorrelation functions for point-process time series
🤖 AI Summary
This paper addresses the challenge of efficiently detecting common temporal patterns—such as periodicity, bursting, and inhibition—in point process time series. We propose a novel autocorrelation plot method specifically designed for point processes. Departing from conventional kernel smoothing, our approach employs a simple binning strategy to directly estimate the autocorrelation function, yielding the first computationally efficient, parameter-free, and theoretically interpretable autocorrelation visualization tool for point process sequences. The estimator is proven to be consistent under mild regularity conditions. Empirically, it accurately identifies diverse canonical temporal patterns in simulated data and demonstrates practical utility in real-world spatiotemporal point processes—including earthquake occurrences and crime incidents—enhancing both interpretability and applicability of pattern discovery.
Technology Category
Application Category
📝 Abstract
This article introduces autocorrelograms for time series of point processes. The method is computationally simple, based on binning rather than smoothing. The ability of the method to detect common time series patterns is shown by simulation, and two examples of application to temporal and spatial point processes series are given.
Problem
Research questions and friction points this paper is trying to address.
Develop autocorrelograms for point-process time series
Use binning instead of smoothing for simplicity
Detect patterns in temporal and spatial series
Innovation
Methods, ideas, or system contributions that make the work stand out.
Autocorrelograms for point-process time series
Computationally simple binning method
Detects common time series patterns
🔎 Similar Papers
No similar papers found.
