Approximate solutions of a general stochastic velocity-jump process subject to discrete-time noisy observations

Arianna Ceccarelli,Alexander P. Browning,Ruth E. Baker

2024-07-05

Abstract:Advances in experimental techniques allow the collection of high-space-and-time resolution data that track individual motile entities over time. This poses the question of how to use these data to efficiently and effectively calibrate motion models. However, typical mathematical models often overlook the inherent aspects of data collection, such as the discreteness and the experimental noise of the measured locations. In this paper, we focus on velocity-jump models suitable to describe single-agent motion in one spatial dimension, characterised by successive Markovian transitions between a finite network of $n$ states, each with a specified velocity and a fixed rate of switching to every other state. Since the problem of finding the exact distributions of discrete-time noisy data is generally intractable, we derive a series of approximations for the data distributions and compare them to in-silico data generated by the models using four example network structures. These comparisons suggest that the approximations are accurate given sufficiently infrequent state switching, or equivalently, a sufficiently high data collection frequency. Moreover, for infrequent switching, the PDFs comparisons highlight the importance of accounting for the correlation between subsequent measured locations, due to the likely permanence in the state visited in the previous measurement. The approximate distributions computed can be used for fast parameter inference and model selection between a range of velocity-jump models using single-agent tracking data.

Data Analysis, Statistics and Probability,Quantitative Methods

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to use the single moving - entity tracking data with high - spatial - and - temporal resolution to calibrate motion models efficiently and effectively. However, existing mathematical models usually overlook some inherent characteristics in the data - collection process, such as the discreteness of measured positions and experimental noise. This paper focuses on the velocity - jump model that describes the motion of a single agent in one - dimensional space. This model consists of continuous Markov transitions in a finite - state network, with each state having a specified velocity and a fixed rate of switching to other states. Since it is usually infeasible to find the exact distribution of discrete - time noisy data, the author derives a series of approximate solutions and compares these approximate solutions with the simulation data generated by the model to verify their accuracy. Moreover, for infrequent state - switching, the comparison of probability density functions (PDFs) emphasizes the importance of considering the correlation between subsequent measured positions, because it is very likely to remain in the state visited in the previous measurement. These approximate distributions can be used for rapid parameter inference and model selection using single - agent tracking data.