Mosquito populations and human social behavior: A spatially explicit agent-based model.

Some mosquitoes are vectors for disease transmission to human populations. Aedes aegypti, the main vector for dengue in Argentina, mainly breeds in artificial containers as it is strongly adapted to urban environments. This highlights the relevance of understanding human social behavior to design successful vector control campaigns. We developed a model of mosquito populations that considers their main biological and behavioral features and incorporates parameters that model human behavior in relation to water container disposal. We performed extensive numerical simulations to study the variability of adult and aquatic mosquito populations when various protocols are applied, changing the effectiveness and frequency of water bucket disposal and the delay in the availability of water containers for breeding. We found an effectiveness threshold value above which it is possible to significantly limit mosquito dispersal. Interestingly, a nonsynchronized discard frequency, more attainable by human populations, was more efficient than a synchronized one to reduce the aquatic mosquito population. Scenarios with random delays in the availability of water containers indicate that it is not decisive to have a fixed time delay for the entire population, which is more realistic as it mimics a wider range of human behaviors. This simple model could help design dengue prevention campaigns aiming at mosquito population control.

Effects of the COVID-19 pandemic in higher education: A data driven analysis for the knowledge acquisition process.

The COVID-19 pandemic abruptly changed the classroom context and presented enormous challenges for all actors in the educational process, who had to overcome multiple difficulties and incorporate new strategies and tools to construct new knowledge. In this work we analyze how student performance was affected, for a particular case of higher education in La Plata, Argentina. We developed an analytical model for the knowledge acquisition process, based on a series of surveys and information on academic performance in both contexts: face-to-face (before the onset of the pandemic) and virtual (during confinement) with 173 students during 2019 and 2020. The information collected allowed us to construct an adequate representation of the process that takes into account the main contributions common to all individuals. We analyzed the significance of the model by means of Artificial Neural Networks and a Multiple Linear Regression Method. We found that the virtual context produced a decrease in motivation to learn. Moreover, the emerging network of contacts built from the interaction between peers reveals different structures in both contexts. In all cases, interaction with teachers turned out to be of the utmost importance in the process of acquiring knowledge. Our results indicate that this process was also strongly influenced by the availability of resources of each student. This reflects the reality of a developing country, which experienced prolonged isolation, giving way to a particular learning context in which we were able to identify key factors that could guide the design of strategies in similar scenarios.

Study of taxes, regulations and inequality using machine learning algorithms.

Genetic machine learning (ML) algorithms to train agents in the Yard-Sale model proved very useful for finding optimal strategies that maximize their wealth. However, the main result indicates that the more significant the fraction of rational agents, the greater the inequality at the collective level. From social and economic viewpoints, this is an undesirable result since high inequality diminishes liquidity and trade. Besides, with very few exceptions, most agents end up with zero wealth, despite the inclusion of rational behaviour. To deal with this situation, here we include a taxation-redistribution mechanism in the ML algorithm. Previous results show that simple regulations can considerably reduce inequality if agents do not change their behaviour. However, when considering rational agents, different types of redistribution favour risk-averse agents, to some extent. Even so, we find that rational agents looking for optimal wealth can always arrive to an optimal risk, compatible with a particular choice of parameters, but increasing inequality. This article is part of the theme issue 'Kinetic exchange models of societies and economies'.

Wealth exchange models and machine learning: Finding optimal risk strategies in multiagent economic systems.

The Yard-Sale Model, a well known wealth exchange model whose observed macroscopic behavior hides many underlying aspects of particular complexity, was studied at the microscopic level. The performance of the agents during the successive transactions allows for the definition of successful or disadvantageous strategies according to the profit they achieve at the end of the process. Optimal strategies were found that maximize the individual wealth of each agent by performing their training through a genetic algorithm. The addition of different levels of rationality given by the amount of available information from their environment showed promising results, at both the microscopic and macroscopic levels. Remarkably, after the training process, the rational agents were able to determine when it would be convenient to interact with their opponents. Additionally, a region of parameters was found for which the distribution of wealth is a power law throughout the whole wealth range. As a general result, the incorporation of rational agents in this type of systems leads to greater inequality at the collective level.

The Knowledge Acquisition Process from a Complex System Perspective: Observations and Models.

We study the knowledge acquisition process in a teaching-learning scenario that takes place within the classroom. We explore two complementary approaches, which include classroom observations and student surveys, and the formulation of theoretical models through the use of statistical physics tools. We develop an analytical model and a set of dynamics agent-based models that allow us to understand global behaviors, as well as to follow individual trajec-tories in the knowledge acquisition process. As a proxy of the final achievements of the students we use their final grade, allowing us to assess the validity of our approach. Our models, supported by observations and surveys, reproduce fairly well the process of acquiring knowledge of the students. This work sheds light on the internal dynamics of the classroom and allows us to understand some global aspects of the teaching-learning process.

Insights About Head Lice Transmission From Field Data and Mathematical Modeling.

Head lice infest millions of school-age children every year, both in developed and developing countries. However, little is known about the number of lice transferred among children during school activities, because direct methods to study this are almost impossible to implement. This issue has been addressed following an indirect method, which consist in collecting data of real infestation from several children groups and using a mathematical model of lice colonies to infer how the infestation observed might have evolved. By determining the events that would most likely lead to infestations as those observed, we find that severe infestations are most likely initiated by a relatively large number of lice transferred at the same moment or within relatively short time spans. In turn, analysis of the data obtained from screenings of the same groups of children a few days apart shows evidence of such transmission events. Interestingly, only children with severe infestations could harbor the lice necessary for this type of transmission. Thus, they play the same role as 'superspreaders' in epidemiology. As part of our experimental study it is also shown that a simple procedure of combing can be very effective to remove all mobile lice, and thus could be used as an effective preventive measure against those severe infestations that are responsible for the spread of pediculosis.

Of lice and math: using models to understand and control populations of head lice.

In this paper we use detailed data about the biology of the head louse (pediculus humanus capitis) to build a model of the evolution of head lice colonies. Using theory and computer simulations, we show that the model can be used to assess the impact of the various strategies usually applied to eradicate head lice, both conscious (treatments) and unconscious (grooming). In the case of treatments, we study the difference in performance that arises when they are applied in systematic and non-systematic ways. Using some reasonable simplifying assumptions (as random mixing of human groups and the same mobility for all life stages of head lice other than eggs) we model the contagion of pediculosis using only one additional parameter. It is shown that this parameter can be tuned to obtain collective infestations whose characteristics are compatible with what is given in the literature on real infestations. We analyze two scenarios: One where group members begin treatment when a similar number of lice are present in each head, and another where there is one individual who starts treatment with a much larger threshold ("superspreader"). For both cases we assess the impact of several collective strategies of treatment.