Exploration–exploitation trade-off features a saltatory search behaviour

Abstract

Searching experiments conducted in different virtual environments over a gender-balanced group of people revealed a gender irrelevant scale-free spread of searching activity on large spatio-temporal scales. We have suggested and solved analytically a simple statistical model of the coherentnoise type describing the exploration-exploitation trade-off in humans ('should I stay' or 'should I go'). The model exhibits a variety of saltatory behaviours, ranging from Lévy flights occurring under uncertainty to Brownian walks performed by a treasure hunter confident of the eventual success.

Consumption and foraging behaviors for common stimulants (nicotine, caffeine)

“…26 In humans this has been likened to the decision as to whether to exploit "here" or explore "there." 27 However, these behaviors may not actually arise from conscious consideration as similar search patterns also 110 occur in lower organisms. 28 Simulations suggest that L evy walk foraging tends to reduce periods of famine.…”

Optimal Lévy-flight foraging in a finite landscape

“…One important improvement to the model is to incorporate an adaptive foraging strategy based on available energy. As a supplement to this work, one can also study other termination conditions to capture more complex internal regularity, memory mechanisms or decision-making of foraging [41,42,43,44,45]. Moreover, one can consider time-variant and heterogenous distribution of targets in the landscape, which can lead to a study of how the foraging strategy adapts to the environmental changes in the presence of cognition and memory.…”

Global optimisation in Hilbert spaces using the survival of the fittest algorithm

“…Note that, the small oscillations in J for a large number of Fourier terms, can be partially explained by effects of noise on the system. In the same figure, we also indicate (dashed yellow line) the fitness corresponding to piecewise linear approximation (8). The figure shows that fitness is much lower with a piecewise linear trajectory (yellow dashed line) when compared to that of the global maximum based on smooth curves (solid red curve).…”

“…1 shown for a progressively increasing number of Fourier terms in (5). The figure also shows the approximation of the trajectories using a piecewise linear function (8), depicted by the dashed lines. Finally, we show other local maximum possible in the system to demonstrate the need for usage of global optimisation.…”

“…Genetic algorithms encode information about an individual via a binary representation, whereas differential evolution considers a non-binary representation [5,6]. The other types of bio-inspired algorithms include swarm intelligence, which is based off interacting searching agents sharing information [7,3], and bacterial foraging, which imitate the social foraging behaviour of Escherichia coli or other microorganisms ( [2]) or even humans ( [8]).…”