Polynomial Random Dynamical Systems with Complete Connections and the Probability of Tending to Infinity

TL;DR

This paper investigates polynomial random dynamical systems with memory on the Riemann sphere, analyzing the probability of orbits tending to infinity and establishing continuity properties of these probabilities under certain conditions.

Contribution

It introduces a framework for systems with state-dependent rules with memory, extending classical models, and proves new continuity and positivity results for escape probabilities.

Findings

Probability of tending to infinity is locally constant on the Fatou set.

If all kernel Julia sets are empty, the escape probability is continuous.

Stationary-averaged escape probabilities form a compact interval and can be positive and nontrivial.

Abstract

We study polynomial random dynamical systems with complete connections on the Riemann sphere. In this framework, the choice of the next polynomial map is governed by a state-dependent rule with memory, extending both i.i.d. random dynamics and non-i.i.d. Markovian models. For each initial state, we define the probability that the random orbit tends to infinity. We prove that it is locally constant on the Fatou set, and that if all kernel Julia sets are empty, then it is continuous on the whole space. We also introduce stationary-averaged escaping probabilities with respect to stationary distributions of the induced state chain. Under the same kernel-emptiness assumption, these averaged probabilities are continuous. In addition, for each point of the Riemann sphere, the set of all possible stationary-averaged values is shown to be a compact interval determined by ergodic stationary distributions. We further give a sufficient condition for the stationary-averaged escaping probability to be everywhere positive and nontrivial. Finally, we provide examples showing RSCC-specific phenomena, including reinforcement-induced discontinuity, recovery of continuity under truncation, and genuinely mixed escaping behavior produced by stationary averaging.