arXiv subjects

Leonardo B. L. Santos

Publications and source records attributed to Leonardo B. L. Santos.

7 recordsLinked to original sources

arXiv2026

Leveraging graph neural networks and mobility data for COVID-19 forecasting

The COVID-19 pandemic has claimed millions of lives, spurring the development of diverse forecasting models. In this context, the true utility of complex spatio-temporal architectures versus simpler temporal baselines remains a subject of debate. Here, we show that structural sparsification of the input graph and temporal granularity are determining factors for the effectiveness of Graph Neural Networks (GNNs). By leveraging human mobility networks in Brazil and China, we address a conflicting scenario in the literature: while standard LSTMs suffice for smooth, monotonic cumulative trends, GNNs significantly outperform baselines when forecasting volatile daily case counts. We show that backbone extraction substantially enhances predictive stability and reduces predictive error by removing negligible connections. Our results indicate that incorporating spatial dependencies is essential for modeling complex dynamics. Specifically, GNN architectures such as GCRN and GCLSTM outperform the LSTM baseline (Nemenyi test, p < 0.05) on datasets from Brazil and China for daily case predictions. Lastly, we frame the problem as a binary classification task to better analyze the dependency between context sizes and prediction horizons.

cs.LG↗

arXiv2023

Precipitation event-based networks: an analysis of the relations between network metrics and meteorological properties

The study of complex systems in nature is essential to understand the interactions between different elements and how they influence one another. Complex network theory is a powerful tool that helps us to analyze these interactions and gain insights into the behavior of such systems. Surprisingly, this theory has been underutilized in the field of weather science, which focuses on the immediate state of the atmosphere. Our research aims to fill this gap by exploring the use of complex network theory in weather science. Specifically, we employ weather radar data to construct event-based geographical networks. By analyzing the relations between meteorological properties and network metrics in these event-based networks, we can gain a better understanding of the behavior of precipitation events. Our findings reveal significant correlations between various meteorological properties and network metrics, shedding light on the underlying mechanisms that govern precipitation events. Through our work, we hope to demonstrate the potential of complex network theory in weather science and inspire further research in this field.

cs.SI↗

arXiv2023

Vulnerability analysis in Complex Networks under a Flood Risk Reduction point of view

The measurement and mapping of transportation network vulnerability to natural hazards constitute subjects of global interest, especially due to climate change, and for a sustainable development agenda. During a flood, some elements of a transportation network can be affected, causing loss of life of people and damage to vehicles, streets/roads, and other logistics services, sometimes with severe economic impacts. The Network Science approach may offer a valuable perspective considering one type of vulnerability related to network type critical infrastructures: the topological vulnerability. The topological vulnerability index associated with an element is defined as the reduction in the network's average efficiency due to the removal of the set of edges related to that element. We present a topological vulnerability index analysis for the highways in the state of Santa Catarina, Brazil, and produce a map considering that index and the areas susceptible to urban floods and landslides. The risk knowledge, combining hazard and vulnerability, is the first pillar of an Early Warning System, and represent an important tool for stakeholders from the transportation sector in a disaster risk reduction agenda.

physics.soc-ph↗

arXiv2022

Beyond the shortest path: the path length index as a distribution

The traditional complex network approach considers only the shortest paths from one node to another, not taking into account several other possible paths. This limitation is significant, for example, in urban mobility studies. In this short report, as the first steps, we present an exhaustive approach to address that problem and show we can go beyond the shortest path, but we do not need to go so far: we present an interactive procedure and an early stop possibility. After presenting some fundamental concepts in graph theory, we presented an analytical solution for the problem of counting the number of possible paths between two nodes in complete graphs, and a depth-limited approach to get all possible paths between each pair of nodes in a general graph (an NP-hard problem). We do not collapse the distribution of path lengths between a pair of nodes into a scalar number, we look at the distribution itself - taking all paths up to a pre-defined path length (considering a truncated distribution), and show the impact of that approach on the most straightforward distance-based graph index: the walk/path length.

cs.DM↗

arXiv2022

Flood risk map from hydrological and mobility data: a case study in São Paulo (Brazil)

Cities increasingly face flood risk primarily due to extensive changes of the natural land cover to built-up areas with impervious surfaces. In urban areas, flood impacts come mainly from road interruption. This paper proposes an urban flood risk map from hydrological and mobility data, considering the megacity of São Paulo, Brazil, as a case study. We estimate the flood susceptibility through the Height Above the Nearest Drainage algorithm; and the potential impact through the exposure and vulnerability components. We aggregate all variables into a regular grid and then classify the cells of each component into three classes: Moderate, High, and Very High. All components, except the flood susceptibility, have few cells in the Very High class. The flood susceptibility component reflects the presence of watercourses, and it has a strong influence on the location of those cells classified as Very High.

physics.soc-ph↗

arXiv2022

Global-threshold and backbone high-resolution weather radar networks are significantly complementary in a watershed

There are several criteria for building up networks from time series related to different points in geographical space. The most used criterion is the Global-Threshold (GT). Using a weather radar dataset, this paper shows that the Backbone (BB) - a local-threshold criterion - generates networks whose geographical configuration is complementary to the GT networks. We compare the results for two well-known similarities measures: the Pearson Correlation (PC) coefficient and the Mutual Information (MI). The extracted backbone network (miBB), whose number of links is the same as the global MI (miGT), has the lowest average shortest path and presents a small-world effect. Regarding the global PC (pcGT) and its corresponding BB network (pcBB), there is a significant linear relationship: $R2=0.77$ with a slope of $1.15$ (p-value $<E-7$) for the pcGT network, and $R2=0.68$ with a slope of $0.76$ (p-value $<E-7$) for the pcBB network. In relation to the MI ones, only the miGT present a high $R2$ ($0.79$, with slope = $1.95$), whereas the miBB has an $R2$ of only $0.20$ ($\text{slope} =0.24$). On the one hand, the GT networks present a sizeable connected component in the central area, close to the main rivers. On the other hand, the BB networks present a few meaningful connected components surrounding the watershed and dominating cells close to the outlet, with significant statistical differences in the altimetry distribution.

cs.SI↗

arXiv2020

Robustness analysis in an inter-cities mobility network: modeling municipal, state and federal initiatives as failures and attacks

Motivated by the challenge related to the COVID-19 epidemic and the seek for optimal containment strategies, we present a robustness analysis into an inter-cities mobility complex network. We abstract municipal initiatives as nodes' failures and the federal actions as targeted attacks. The geo(graphs) approach is applied to visualize the geographical graph and produce maps of topological indexes, such as degree and vulnerability. A Brazilian data of 2016 is considered a case study, with more than five thousand cities and twenty-seven states. Based on the Network Robustness index, we show that the most efficient attack strategy shifts from a topological degree-based, for the all cities network, to a topological vulnerability-based, for a network considering the Brazilian States as nodes. Moreover, our results reveal that individual municipalities' actions do not cause a high impact on mobility restrain since they tend to be punctual and disconnected to the country scene as a whole. Oppositely, the coordinated isolation of specific cities is key to detach entire network areas and thus prevent a spreading process to prevail.

physics.soc-ph↗