1. Learning physical properties of liquid crystals with deep convolutional neural networksHigor Y. D. Sigaki, Ervin K. Lenzi, Rafael S. Zola, Matjaž Perc, Haroldo V. Ribeiro, 2020, original scientific article Abstract: Machine learning algorithms have been available since the 1990s, but it is much more recently that they have come into use also in the physical sciences. While these algorithms have already proven to be useful in uncovering new properties of materials and in simplifying experimental protocols, their usage in liquid crystals research is still limited. This is surprising because optical imaging techniques are often applied in this line of research, and it is precisely with images that machine learning algorithms have achieved major breakthroughs in recent years. Here we use convolutional neural networks to probe several properties of liquid crystals directly from their optical images and without using manual feature engineering. By optimizing simple architectures, we fnd that convolutional neural networks can predict physical properties of liquid crystals with exceptional accuracy. We show that these deep neural networks identify liquid crystal phases and predict the order parameter of simulated nematic liquid crystals almost perfectly. We also show that convolutional neural networks identify the pitch length of simulated samples of cholesteric liquid crystals and the sample temperature of an experimental liquid crystal with very high precision.
Keywords: liquid crystal, neural network, artificial intelligence, soft matter
Published in DKUM: 20.11.2024; Views: 0; Downloads: 3
 Full text (1,94 MB)
This document has many files! More... |
2. Interplay between particle trapping and heterogeneity in anomalous diffusionHaroldo V. Ribeiro, Angel A. Tateishi, Ervin K. Lenzi, Richard L. Magin, Matjaž Perc, 2023, original scientific article Abstract: Heterogeneous media diffusion is often described using position-dependent diffusion coefficients and estimated indirectly through mean squared displacement in experiments. This approach may overlook other mechanisms and their interaction with position-dependent diffusion, potentially leading to erroneous conclusions. Here, we introduce a hybrid diffusion model that merges a position-dependent diffusion coefficient with the trapping mechanism of the comb model. We derive exact solutions for position distributions and mean squared displacements, validated through simulations of Langevin equations. Our model shows that the trapping mechanism attenuates the impact of media heterogeneity. Superdiffusion occurs when the position-dependent coefficient increases superlinearly, while subdiffusion occurs for sublinear and inverse power-law relations. This nontrivial interplay between heterogeneity and state-independent mechanisms also leads to anomalous yet Brownian, and non-Brownian yet Gaussian regimes. These findings emphasize the need for cautious interpretations of experiments and highlight the limitations of relying solely on mean squared displacements or position distributions for diffusion characterization.
Keywords: particle trapping, heterogeneity, diffusion, statistical physics
Published in DKUM: 11.09.2024; Views: 38; Downloads: 4
Full text (3,73 MB)
This document has many files! More... |
3. Deep learning criminal networksHaroldo V. Ribeiro, Diego D. Lopes, Arthur A. B. Pessa, Alvaro F. Martins, Bruno R. da Cunha, Sebastián Gonçalves, Ervin K. Lenzi, Quentin S. Hanley, Matjaž Perc, 2023, original scientific article Abstract: Recent advances in deep learning methods have enabled researchers to develop and apply algorithms for the analysis and modeling of complex networks. These advances have sparked a surge of interest at the interface between network science and machine learning. Despite this, the use of machine learning methods to investigate criminal networks remains surprisingly scarce. Here, we explore the potential of graph convolutional networks to learn patterns among networked criminals and to predict various properties of criminal networks. Using empirical data from political corruption, criminal police intelligence, and criminal financial networks, we develop a series of deep learning models based on the GraphSAGE framework that are able to recover missing criminal partnerships, distinguish among types of associations, predict the amount of money exchanged among criminal agents, and even anticipate partnerships and recidivism of criminals during the growth dynamics of corruption networks, all with impressive accuracy. Our deep learning models significantly outperform previous shallow learning approaches and produce high-quality embeddings for node and edge properties. Moreover, these models inherit all the advantages of the GraphSAGE framework, including the generalization to unseen nodes and scaling up to large graph structures.
Keywords: organized crime, complexity, crime prediction, GraphSAGE
Published in DKUM: 20.06.2024; Views: 236; Downloads: 18
Full text (2,36 MB)
This document has many files! More... |
4. Machine learning partners in criminal networksDiego D. Lopes, Bruno R. da Cunha, Alvaro F. Martins, Sebastián Gonçalves, Ervin K. Lenzi, Quentin S. Hanley, Matjaž Perc, Haroldo V. Ribeiro, 2022, original scientific article Abstract: Recent research has shown that criminal networks have complex organizational structures, but whether this can be used to predict static and dynamic properties of criminal networks remains little explored. Here, by combining graph representation learning and machine learning methods, we show that structural properties of political corruption, police intelligence, and money laundering networks can be used to recover missing criminal partnerships, distinguish among diferent types of criminal and legal associations, as well as predict the total amount of money exchanged among criminal agents, all with outstanding accuracy. We also show that our approach can anticipate future criminal associations during the dynamic growth of corruption networks with signifcant accuracy. Thus, similar to evidence found at crime scenes, we conclude that structural patterns of criminal networks carry crucial information about illegal activities, which allows machine learning methods to predict missing information and even anticipate future criminal behavior.
Keywords: machine learning, crime, network, social physics
Published in DKUM: 28.05.2024; Views: 655; Downloads: 12
Full text (2,42 MB)
This document has many files! More... |
