Search

Additional file 4: Figure S4. Dynamics of relative abundance.

Additional file 12: Figure S12. Comparison of nonlinear and linear forecasting approaches.

Additional file 13: Figure S13. Candidates of signal indices for anticipating abrupt community changes.

Additional file 8: Figure S8. Examples of population-level forecasting results.

Additional file 1: Figure S1. Experimental setting and microbiome data formats.

Additional file 3: Figure S3. Dynamics of calibrated abundance.

Additional file 7: Figure S7. Histogram of optimal embedding dimension (E) of ASVs.

Additional file 11: Figure S11. Distribution of prediction error in the community-level forecasting.

Additional file 9: Figure S9. Dependence of population-level forecasting results on reference database size.

Additional file 5: Figure S5. Distribution of stable states on the energy landscapes.

Additional file 15: Figure S15. ROC analysis of diagnostic performance.

Additional file 10: Figure S10. Comparison of predicted and observed community structure (seven-day-ahead forecasting).

Additional file 2: Figure S2. Dynamics of calibrated abundance without interpolation.

Scripts and figures for the paper.

Supplemental material, sj-pdf-1-jbr-10.1177_07487304211002934 for Individual Ants Do Not Show Activity-Rest Rhythms in Nest Conditions by Haruna Fujioka, Masato S. Abe and Yasukazu Okada in Journal of Biological Rhythms

Division of labour is the key factor leading to higher-order systems such as cooperative animal groups. How division of labour is achieved without a central control is one of the most fascinating questions in behavioural ecology and complex systems science. Social insects are one of the best examples of complex self-organised systems through local interactions. However, it is difficult to comprehensively understand division of labour due to the chronological and individual variation in behaviours, and the differences between social environments. Thus, it is imperative to quantify individual behaviours and integrate them into colony levels. Here, we demonstrate that network analyses on individual-behaviour relationships can be valid for characterising the task allocation patterns. We recorded the behaviours of all individuals with verified age in colonies of monomorphic ant and analysed the individual-behaviour relationship at individual, sub-network, and network-levels. We detected the “task groups” including subsets of individuals and behaviour, and the patterns of task groups were consistent throughout the day. We also found a relationship between the age and the characteristics of individuals within the network. These findings suggest that bipartite network analyses can untangle the complex task allocation patterns and provide novel insights into understanding how division of labour is achieved.

A bstract In biological movements, a special class of random walks, the so-called Lévy walk, has been widely observed. It has been shown that Lévy walks outperform normal random walks in terms of search efficiency because of rare, long, straight movements among short steps, and thus the cause of its prevalence is considered to be a consequence of natural selection. However, recent findings suggest that Lévy walks might also be the result of other mechanisms. This suggests that its origins depend on ecological and physical conditions. Although it is crucial to explore the mechanisms of producing Lévy walks by considering various conditions, the generative mechanisms of Lévy walks based on internal states with nonlinear dynamics are less understood. Here, we develop a generative model composed of simple deterministic nonlinear systems and show that the Lévy walk of an autonomous agent can emerge near the critical point, which is between the synchronous and asynchronous states. Furthermore, we show that Lévy walks have maximized sensitivity to a perturbation and can produce diverse behavior, which might be another advantage of Lévy walks. Our results suggest that the commonly observed Lévy walk can be derived from the critical point hypothesis in which biological systems sitting in a critical point between order and disorder receive benefits. This provides new insights into understandings of how and why Lévy walks are prevalent in biological systems.

Table S1 and Figures S1-S4

We present a practical health-theme machine learning (ML) application concerning `AI for social good' domain for `Producing Good Outcomes' track. In particular, the solution is concerning the problem of a potential elderly adult dementia onset prediction in aging societies. The paper discusses our attempt and encouraging preliminary study results of behavioral responses analysis in a working memory-based emotional evaluation experiment. We focus on the development of digital biomarkers for dementia progress detection and monitoring. We present a behavioral data collection concept for a subsequent AI-based application together with a range of regression encouraging results of Montreal Cognitive Assessment (MoCA) scores in the leave-one-subject-out cross-validation setup. The regressor input variables include experimental subject's emotional valence and arousal recognition responses, as well as reaction times, together with self-reported education levels and ages, obtained from a group of twenty older adults taking part in the reported data collection project. The presented results showcase the potential social benefits of artificial intelligence application for elderly and establish a step forward to develop ML approaches, for the subsequent application of simple behavioral objective testing for dementia onset diagnostics replacing subjective MoCA., Comment: 4 pages, 2 figures, accepted for a presentation at AI for Social Good workshop at NeurIPS (2019), Vancouver, Canada

Figs S1-S7 and Table S1

Supplementary Methods and Figs. S1 S2 S3