Abstract:The pair-contact process with diffusion (PCPD), a generalized model of the ordinary pair-contact process (PCP) without diffusion, exhibits a continuous absorbing phase transition. Unlike the PCP, whose nature of phase transition is clearly classified into the directed percolation (DP) universality class, the model of PCPD has been controversially discussed since its infancy. To our best knowledge, there is so far no consensus on whether the phase transition of the PCPD falls into the unknown university classes or else conveys a new kind of non-equilibrium phase transition. In this paper, both unsupervised and supervised learning are employed to study the PCPD with scrutiny. Firstly, two unsupervised learning methods, principal component analysis (PCA) and autoencoder, are taken. Our results show that both methods can cluster the original configurations of the model and provide reasonable estimates of thresholds. Therefore, no matter whether the non-equilibrium lattice model is a random process of unitary (for instance the DP) or binary (for instance the PCP), or whether it contains the diffusion motion of particles, unsupervised leaning can capture the essential, hidden information. Beyond that, supervised learning is also applied to learning the PCPD at different diffusion rates. We proposed a more accurate numerical method to determine the spatial correlation exponent $\nu_{\perp}$, which, to a large degree, avoids the uncertainty of data collapses through naked eyes. Our extensive calculations reveal that $\nu_{\perp}$ of PCPD depends continuously on the diffusion rate $D$, which supports the viewpoint that the PCPD may lead to a new type of absorbing phase transition.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is about the non - equilibrium phase - transition characteristics of the Pair - Contact Process with Diffusion (PCPD). Specifically, the paper aims to explore whether PCPD belongs to the known universality class or whether it exhibits a new type of non - equilibrium phase - transition. PCPD is a generalized ordinary Pair - Contact Process (PCP) model. The latter exhibits clear phase - transition characteristics in the absence of diffusion and is classified into the Directed Percolation (DP) universality class. However, for the PCPD model that includes diffusion, its phase - transition nature has been controversial and no consensus has been reached yet. ### Main research objectives of the paper: 1. **Study the PCPD model using machine - learning techniques**: The paper adopts unsupervised learning and supervised learning methods to deeply analyze the PCPD model. Unsupervised learning methods include Principal Component Analysis (PCA) and Autoencoder, which are used to cluster the original configurations of the model and provide a reasonable estimate of the threshold. Supervised learning is used to learn the PCPD model at different diffusion rates and proposes a more accurate numerical method to determine the spatial correlation index $\nu_\perp$. 2. **Verify the phase - transition nature of the PCPD model**: Through the above - mentioned machine - learning methods, the paper hopes to verify the phase - transition point of the PCPD model and explore whether it belongs to the known universality class or exhibits new non - equilibrium phase - transition characteristics. ### Main research methods: - **Unsupervised learning**: - **Principal Component Analysis (PCA)**: Used for dimension reduction and data visualization to extract the hidden information of the model. - **Autoencoder**: Used for processing nonlinear data, performing image recognition and data compression. - **Supervised learning**: - **Classification and regression**: Used for identifying the phase - transition point and predicting the phase - transition type, trained and tested with data generated by Monte Carlo simulation. ### Research background: - **Non - equilibrium phase - transition**: Different from equilibrium phase - transition, non - equilibrium phase - transition involves changes in the time dimension, and the system state and properties change over time. The PCPD model is a binary random reaction process, and when the diffusion rate is zero, it degenerates into an ordinary Pair - Contact Process (PCP). PCP has been proven to belong to the Directed Percolation (DP) universality class, but the phase - transition nature of PCPD is still controversial. - **Application of machine learning in phase - transition research**: In recent years, machine - learning techniques have made significant progress in the research of equilibrium and non - equilibrium phase - transitions. Unsupervised learning methods such as PCA and Autoencoder are excellent in extracting key features of complex data, while supervised learning methods have advantages in identifying phase - transition points and predicting phase - transition types. ### Research significance: - **Theoretical contribution**: Through machine - learning techniques, the paper provides a new perspective for understanding the phase - transition nature of the PCPD model, which helps to solve the long - standing controversy about the PCPD universality class. - **Methodological innovation**: Demonstrates the application potential of machine learning in non - equilibrium phase - transition research and provides methodological support for future similar research. In conclusion, this paper conducts a comprehensive study of the PCPD model by combining unsupervised and supervised learning methods, aiming to reveal its phase - transition characteristics and universality - class membership.

Machine learning of pair-contact process with diffusion

Machine learning of pair-contact process with diffusion

A coupled two-species model for the pair contact process with diffusion

Supervised and unsupervised learning of directed percolation

Machine learning of (1+1)-dimensional directed percolation based on raw and shuffled configurations

Transfer learning of phase transitions in percolation and directed percolation

Machine Learning Percolation Model

Machine learning for percolation utilizing auxiliary Ising variables

Supervised and unsupervised learning of (1+1)-dimensional even-offspring branching annihilating random walks

Machine Learning of Phase Transitions in the Percolation and XY Models

Learning phase transitions by siamese neural network

Machine Learning Dynamical Phase Transitions in Complex Networks.

Universality classes of absorbing phase transitions in generic branching-annihilating particle systems

Machine Learning Based on Wave and Diffusion Physical Systems

Contact processes in crowded environments

Detecting phase transitions based on siamese neural network

Exploring percolation phase transition in the three-dimensional Ising model with machine learning

Principal component analysis of absorbing state phase transitions

Characterisation of dislocation patterning behaviour with a continuum dislocation dynamics model on two parallel slip planes equipped with a deep neural network resolving local microstructures

Determination of the critical points for systems of directed percolation class using machine learning