Amoebae for clustering: a bio-inspired cellular automata method for data classification

Amaury Saint-Jore (1) , Nazim A. Fatès (1) , Emmanuel Jeandel (1)

1 MOCQUA - Designing the Future of Computational Models

We present a bio-inspired mechanism for data clustering. Our method uses amoebae which evolve according to cellular automata rules: they contain the data to be processed and emit reaction-diffusion waves at random times. The waves transmit the information across the lattice and causes other amoebae to react, by being attracted or repulsed. The local reactions produce small homogeneous groups which progressively merge and realise the clustering at a larger scale. Despite the simplicity of the local rules, interesting complex behaviour occur, which make the model robust to various changes of its settings. We evaluate this prototype with a simple task: the separation of two groups of integer values distributed according to Gaussian laws.

Mots clés: Bio-inspired models; Cellular automata; Data clustering; Discrete dynamical systems; Collective intelligence

Domaines: Automates cellulaires et gaz sur réseau [nlin.CG]

Amaury Saint-Jore, Nazim A. Fatès, Emmanuel Jeandel. Amoebae for clustering: a bio-inspired cellular automata method for data classification. Automata and Complexity, 42, Springer, Cham, pp.417-432, 2022, 978-3-030-92553-6. ⟨10.1007/978-3-030-92551-2_23⟩. ⟨hal-02973830⟩