Joost H. van der Linden

Bio: Joost H. van der Linden is an academic researcher from University of Melbourne. The author has contributed to research in topics: Discrete element method & Glycemic. The author has an hindex of 6, co-authored 12 publications receiving 113 citations.

Machine learning framework for analysis of transport through complex networks in porous, granular media: A focus on permeability.

Abstract: We present a data-driven framework to study the relationship between fluid flow at the macroscale and the internal pore structure, across the micro- and mesoscales, in porous, granular media. Sphere packings with varying particle size distribution and confining pressure are generated using the discrete element method. For each sample, a finite element analysis of the fluid flow is performed to compute the permeability. We construct a pore network and a particle contact network to quantify the connectivity of the pores and particles across the mesoscopic spatial scales. Machine learning techniques for feature selection are employed to identify sets of microstructural properties and multiscale complex network features that optimally characterize permeability. We find a linear correlation (in log-log scale) between permeability and the average closeness centrality of the weighted pore network. With the pore network links weighted by the local conductance, the average closeness centrality represents a multiscale measure of efficiency of flow through the pore network in terms of the mean geodesic distance (or shortest path) between all pore bodies in the pore network. Specifically, this study objectively quantifies a hypothesized link between high permeability and efficient shortest paths that thread through relatively large pore bodies connected to each other by high conductance pore throats, embodying connectivity and pore structure.

Real-Time Continuous Glucose Monitoring During the Coronavirus Disease 2019 Pandemic and Its Impact on Time in Range.

Abstract: Background: The coronavirus disease 2019 (COVID-19) pandemic disrupted the lives of people with diabetes. Use of real-time continuous glucose monitoring (rtCGM) helped manage diabetes effectively. Some of these disruptions may be reflected in population-scale changes to metrics of glycemic control, such as time-in-range (TIR). Methods: We examined data from 65,067 U.S.-based users of the G6 rtCGM System (Dexcom, Inc., San Diego, CA) who had uploaded data before and during the COVID-19 pandemic. Users associated with three counties that included the cities of Los Angeles, Chicago, and New York or with five regions designated by the Centers for Disease Control and Prevention (CDC) were compared. Public data were used to associate regions with prepandemic and intrapandemic glycemic parameters, COVID-19 mortality, and median household income. Results: Compared with an 8-week prepandemic interval before stay-at-home orders (January 6, 2020, to March 1, 2020), overall mean (standard deviation) TIR improved from 59.0 (20.1)% to 61.0 (20.4)% during the early pandemic period (April 20, 2020 to June 14, 2020, P < 0.001). TIR improvements were noted in all three counties and in all five CDC-designated regions. Higher COVID-19 mortality was associated with higher proportions of individuals experiencing TIR improvements of ≥5 percentage points. Users in economically wealthier zip codes had higher pre- and intrapandemic TIR values and greater relative improvements in TIR. TIR and pandemic-related improvements in TIR varied across CDC-designated regions. Conclusions: Population-level rtCGM data may be used to monitor changes in glycemic control with temporal and geographic specificity. The COVID-19 pandemic is associated with improvements in TIR, which were not evenly distributed across the United States.

Preferential flow pathways in a deforming granular material: self-organization into functional groups for optimized global transport

Abstract: Existing definitions of where and why preferential flow in porous media occurs, or will occur, assume a priori knowledge of the fluid flow and do not fully account for the connectivity of available flow paths in the system. Here we propose a method for identifying preferential pathways through a flow network, given its topology and finite link capacities. Using data from a deforming granular medium, we show that the preferential pathways form a set of percolating pathways that is optimized for global transport of interstitial pore fluid in alignment with the applied pressure gradient. Two functional subgroups emerge. The primary subgroup comprises the main arterial paths that transmit the greatest flow through shortest possible routes. The secondary subgroup comprises inter- and intra-connecting bridges that connect the primary paths, provide alternative flow routes, and distribute flow through the system to maximize throughput. We examine the multiscale relationship between functionality and subgroup structure as the sample dilates in the lead up to the failure regime where the global volume then remains constant. Preferential flow pathways chain together large, well-connected pores, reminiscent of force chain structures that transmit the majority of the load in the solid grain phase.

IEEE transactions on pattern analysis and machine intelligence

Abstract: In the real world, a realistic setting for computer vision or multimedia recognition problems is that we have some classes containing lots of training data and many classes contain a small amount of training data. Therefore, how to use frequent classes to help learning rare classes for which it is harder to collect the training data is an open question. Learning with Shared Information is an emerging topic in machine learning, computer vision and multimedia analysis. There are different level of components that can be shared during concept modeling and machine learning stages, such as sharing generic object parts, sharing attributes, sharing transformations, sharing regularization parameters and sharing training examples, etc. Regarding the specific methods, multi-task learning, transfer learning and deep learning can be seen as using different strategies to share information. These learning with shared information methods are very effective in solving real-world large-scale problems. This special issue aims at gathering the recent advances in learning with shared information methods and their applications in computer vision and multimedia analysis. Both state-of-the-art works, as well as literature reviews, are welcome for submission. Papers addressing interesting real-world computer vision and multimedia applications are especially encouraged. Topics of interest include, but are not limited to: • Multi-task learning or transfer learning for large-scale computer vision and multimedia analysis • Deep learning for large-scale computer vision and multimedia analysis • Multi-modal approach for large-scale computer vision and multimedia analysis • Different sharing strategies, e.g., sharing generic object parts, sharing attributes, sharing transformations, sharing regularization parameters and sharing training examples, • Real-world computer vision and multimedia applications based on learning with shared information, e.g., event detection, object recognition, object detection, action recognition, human head pose estimation, object tracking, location-based services, semantic indexing. • New datasets and metrics to evaluate the benefit of the proposed sharing ability for the specific computer vision or multimedia problem. • Survey papers regarding the topic of learning with shared information. Authors who are unsure whether their planned submission is in scope may contact the guest editors prior to the submission deadline with an abstract, in order to receive feedback.

Network analysis of particles and grains

Abstract: The arrangements of particles and forces in granular materials have a complex organization on multiple spatial scales that ranges from local structures to mesoscale and system-wide ones. This multiscale organization can affect how a material responds or reconfigures when exposed to external perturbations or loading. The theoretical study of particle-level, force-chain, domain, and bulk properties requires the development and application of appropriate physical, mathematical, statistical, and computational frameworks. Traditionally, granular materials have been investigated using particulate or continuum models, each of which tends to be implicitly agnostic to multiscale organization. Recently, tools from network science have emerged as powerful approaches for probing and characterizing heterogeneous architectures across different scales in complex systems, and a diverse set of methods have yielded fascinating insights into granular materials. In this paper, we review work on network-based approaches to studying granular matter and explore the potential of such frameworks to provide a useful description of these systems and to enhance understanding of their underlying physics. We also outline a few open questions and highlight particularly promising future directions in the analysis and design of granular matter and other kinds of material networks.

Mechanics of granular materials : an introduction

Abstract: This textbook compiles reports written by about 35 internationally recognized authorities, and covers a range of interests for geotechnical engineers. Topics include: fundamentals for mechanics of granular materials; continuum theory of granular materials; and discrete element approaches.

Linking Morphology of Porous Media to Their Macroscopic Permeability by Deep Learning

Abstract: Flow, transport, mechanical, and fracture properties of porous media depend on their morphology and are usually estimated by experimental and/or computational methods. The precision of the computational approaches depends on the accuracy of the model that represents the morphology. If high accuracy is required, the computations and even experiments can be quite time-consuming. At the same time, linking the morphology directly to the permeability, as well as other important flow and transport properties, has been a long-standing problem. In this paper, we develop a new network that utilizes a deep learning (DL) algorithm to link the morphology of porous media to their permeability. The network is neither a purely traditional artificial neural network (ANN), nor is it a purely DL algorithm, but, rather, it is a hybrid of both. The input data include three-dimensional images of sandstones, hundreds of their stochastic realizations generated by a reconstruction method, and synthetic unconsolidated porous media produced by a Boolean method. To develop the network, we first extract important features of the images using a DL algorithm and then feed them to an ANN to estimate the permeabilities. We demonstrate that the network is successfully trained, such that it can develop accurate correlations between the morphology of porous media and their effective permeability. The high accuracy of the network is demonstrated by its predictions for the permeability of a variety of porous media.

Network Analysis of Particles and Grains

Abstract: The arrangements of particles and forces in granular materials have a complex organization on multiple spatial scales that ranges from local structures to mesoscale and system-wide ones. This multiscale organization can affect how a material responds or reconfigures when exposed to external perturbations or loading. The theoretical study of particle-level, force-chain, domain, and bulk properties requires the development and application of appropriate physical, mathematical, statistical, and computational frameworks. Traditionally, granular materials have been investigated using particulate or continuum models, each of which tends to be implicitly agnostic to multiscale organization. Recently, tools from network science have emerged as powerful approaches for probing and characterizing heterogeneous architectures across different scales in complex systems, and a diverse set of methods have yielded fascinating insights into granular materials. In this paper, we review work on network-based approaches to studying granular matter and explore the potential of such frameworks to provide a useful description of these systems and to enhance understanding of their underlying physics. We also outline a few open questions and highlight particularly promising future directions in the analysis and design of granular matter and other kinds of material networks.