Search

Research on brain-like models with bio-rationality promotes the development of artificial intelligence. However, the topology of a brain-like model still lacks bio-rationality. Bio-brains have self-adaptive robustness. The purpose of this paper is to investigate a more bio-rational brain-like model verified by anti-damage capabilities. Thus, this paper proposes a new spiking neural network (SNN) constrained by a functional brain network based on human-brain functional magnetic resonance imaging (fMRI-SNN). Then, we investigate the anti-damage capabilities of our fMRI-SNN, and discuss the mechanism of these anti-damage capabilities. Our results indicate the following: (i) The fMRI-SNN has anti-damage capabilities under random attacks evaluated by two anti-damage indicators. Based on the relevance analysis, our discussion implies that the intrinsic element of the anti-damage capabilities is the synaptic plasticity. (ii) The anti-damage capabilities of our fMRI-SNN outperform those of the scale-free SNN and the small-world SNN. Our discussion on dynamic topological characteristics further implies that the network topology is an element that impacts the performance level of the anti-damage capabilities. • A more bio-rational brain-like model (fMRI-SNN) is proposed. • The performance of fMRI-SNN is verified by anti-damage capabilities. • The anti-damage mechanism of fMRI-SNN is discussed. [ABSTRACT FROM AUTHOR]

Anticipating price crises in the market of agri-commodities is critical to guarantee both the sustainability of the food system and to ensure food security. However, this is not an easy task, since the problem implies analyzing small and very volatile time series, which are highly influenced by external factors. In this paper, we show that suitable reservoir computing algorithms can be developed that outperform traditional approaches, by reducing the Mean Absolute Error and, more importantly, increasing the Market Direction Accuracy. For this purpose, the applicability of five variants of such method to forecast this market is explored, and their performance evaluated by comparing the results with those obtained with the standard LSTM and SARIMA benchmarks. We conclude that decomposing the time series and modeling each component with a separate RC is essential to successfully anticipate price trends, and that this method works even in the complex changing temporal scenario of the Covid-19 pandemic, when part of the data were collected. • We tackle current food loss problem. • Explore predictability in agri-food market. • Successfully anticipate sudden price crisis. • Use the recent ML Reservoir Computing method. • We outperform traditional benchmark models: SARIMA and LSTM. [ABSTRACT FROM AUTHOR]

One of the key tasks in the economy is forecasting the economic agents' expectations of the future values of economic variables using mathematical models. The behavior of mathematical models can be irregular, including chaotic, which reduces their predictive power. In this paper, we study the regimes of behavior of two economic models and identify irregular dynamics in them. Using these models as an example, we demonstrate the effectiveness of evolutionary algorithms and the continuous deep Q-learning method in combination with Pyragas control method for deriving a control action that stabilizes unstable periodic trajectories and suppresses chaotic dynamics. We compare qualitative and quantitative characteristics of the model's dynamics before and after applying control and verify the obtained results by numerical simulation. Proposed approach can improve the reliability of forecasting and tuning of the economic mechanism to achieve maximum decision-making efficiency. • Irregular dynamics in an overlapping generation model is studied. • Chaotic regime is forecasted and stabilized in the spatio-temporal pricing model. • Control variables are synthesized in the Pyragas form. • Evolutionary algorithms are used to find parameters stabilizing irregular dynamics. • Continuous deep Q-learning method is effective for suppressing chaos in the models. [ABSTRACT FROM AUTHOR]

Developing new self-oscillating systems would greatly expand their applications in intelligent machines, advanced robotics, and biomedical devices, due to their advantages of directly harvesting ambient energy and self-control. In this paper, we creatively proposed a light-fueled self-oscillating liquid crystal elastomer (LCE) balloon with self-shading coating, which is capable of self-oscillating under steady ambient illumination. Based on the well-established dynamic LCE model and the ideal gas model, the governing equation of the free-standing LCE balloon is formulated, and dynamics of the self-oscillation is studied theoretically. Numerical calculations show that the balloon always evolves into two kinds of motion modes: static mode and oscillation mode. The mechanism of the self-oscillation is elucidated by the self-shading effect of the opaque power coating, and the conditions for triggering the self-oscillation are further obtained numerically. In addition, the effects of system parameters on amplitude, frequency and equilibrium position are also investigated. Amplitude and frequency are mainly affected by light intensity, damping coefficient, mass density and contraction coefficient, while are independent on initial velocity. It is anticipated that the self-oscillating balloon constructed in this paper would be of benefit in autonomous, self-sustained machines and devices with the core feature of photo-mechanical transduction. [ABSTRACT FROM AUTHOR]

• The FitzHugh-Nagumo neuronal circuit with piecewise memristor is investigated through the theory of switching flow. • The analyticalconditionsare investigatedto understand the switching motionson the boundary. • Thebifurcation diagrams are studied to reveal the hidden extreme multistability. • The parameter mappings and basins of attractor arecarried out to express the coexistingattractors. Memristor, as a nonlinear electronic component, has good switching characteristics, which makes it has excellent application in the field of artificial intelligence. The study for discontinuous dynamics of memristor-based neural network can best reflect the boundary effect of memristors and the complex nonlinear behaviors of system. In this paper, the non-autonomous FitzHugh-Nagumo neuronal circuit with piecewise memristor is investigated in flux-charge domain through the theory of switching flow. The sufficient and necessary conditions of switching motions, such as passable and grazing on the boundary, are developed to understand the switching mechanism. The switching bifurcation diagrams with changing the system parameters and initial conditions are studied to reveal the hidden extreme multistability and coexisting attractors. The parameter mappings and basins of attractor are also carried out to express the coexistences of periodic orbits and chaos with different mapping structures. The simulation results show that the switching and grazing motion are verified the effectiveness the analysis conditions. [ABSTRACT FROM AUTHOR]