Reduced Models of Gene Regulatory Networks: Visualising Multi-Modal Landscapes (To appear)
Alyahya, Khulood . 2021
We present a prototypical multi-modal optimisation problem from the systems biology domain -- tuning the kinetic parameters of a reduced-order gene regulatory network (GRN) model to obtain optimal fits to gene expression timeseries. After introducing the problem, the chapter then illustrates different fitness landscapes of the GRN parameter fitting problem using various statistical plots of landscape features, along with local optima networks (LONs) -- graphs representing local optima (modes), their basin sizes and connectivity across the landscape. As well as presenting problem owners with a putative list of modal solutions from which to verify and select a design from a multi-modal optimisation process, it is often useful to present a characterisation of the search landscape itself. This can provide insight into the domain, and may guide, for instance, problem reformulations, or the final mode selection based on broader features than simply mode performance (e.g. basin size if robustness of modes is a concern)
The Artificial Bee Colony (ABC) is a recently introduced swarm intelligence algorithm for optimization, that has previously been applied successfully to the training of neural networks.
The Artificial Bee Colony (ABC) is a swarm intelligence algorithm for optimization that has previously been applied to the training of neural networks.
This paper empirically studies basic properties of the fitness landscape of random instances of number partitioning problem, with a focus on how these properties change with the phase transition.…