An official website of the United States government
Abstract While the human body has many different examples of perfusable structures with complex geometries, biofabrication methods to replicate this complexity are still lacking. Specifically, the fabrication of self‐supporting, branched networks with multiple channel diameters is particularly challenging. Herein, the Gelation of Uniform Interfacial Diffusant in Embedded 3D Printing (GUIDE‐3DP) approach for constructing perfusable networks of interconnected channels with precise control over branching geometries and vessel sizes is presented. To achieve user‐specified channel dimensions, this technique leverages the predictable diffusion of cross‐linking reaction‐initiators released from sacrificial inks printed within a hydrogel precursor. The versatility of GUIDE‐3DP to be adapted for use with diverse physicochemical cross‐linking mechanisms is demonstrated by designing seven printable material systems. Importantly, GUIDE‐3DP allows for the independent tunability of both the inner and outer diameters of the printed channels and the ability to fabricate seamless junctions at branch points. This 3D bioprinting platform is uniquely suited for fabricating lumenized structures with complex shapes characteristic of multiple hollow vessels throughout the body. As an exemplary application, the fabrication of vasculature‐like networks lined with endothelial cells is demonstrated. GUIDE‐3DP represents an important advance toward the fabrication of self‐supporting, physiologically relevant networks with intricate and perfusable geometries.
more »
« less
This content will become publicly available on August 1, 2026
Machine learning approach for the flexural strength of 3D ‐printed fiber‐reinforced concrete based on the meta‐heuristic algorithm
Abstract The increasing demand for concrete in construction presents challenges such as pollution, high energy consumption, and complex structural requirements. Three‐dimensional printing (3DP) offers a promising solution by eliminating formwork, reducing waste, and enabling intricate geometries. Predicting the strength of 3D‐printed fiber‐reinforced concrete (3DP‐FRC) remains challenging due to the nonlinear nature of neural networks and uncertainty in optimizing key parameters. In this study, we developed machine learning models using five metaheuristic algorithms—arithmetic optimization algorithm, African Vulture Optimization Algorithm, flow direction algorithm, generalized normal distribution optimization, and Mountain Gazelle Optimizer—to optimize the weights and biases in a feed‐forward backpropagation network. Among all the algorithms, MGO demonstrated the best performance. To address data limitations, a data augmentation method combining Kernel density estimation and Wasserstein generative adversarial networks is employed. Sensitivity analysis using SHapley Additive exPlanations (SHAP) identifies the most influential input parameters. The proposed MGO‐ANN model enhances predictive accuracy, reducing the need for extensive laboratory testing. Additionally, a user‐friendly graphical user interface is developed to facilitate practical applications in estimating 3DP‐FRC flexural strength.
more »
« less
- Award ID(s):
- 2310807
- PAR ID:
- 10635268
- Publisher / Repository:
- Wiley
- Date Published:
- Journal Name:
- Structural Concrete
- Volume:
- 26
- Issue:
- 4
- ISSN:
- 1464-4177
- Page Range / eLocation ID:
- 4103 to 4142
- Subject(s) / Keyword(s):
- Structural investigation machine learning
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Machine learning (ML)-based data-driven methods have promoted the progress of modeling in many engineering domains. These methods can achieve high prediction and generalization performance for large, high-quality datasets. However, ML methods can yield biased predictions if the observed data (i.e., response variable y) are corrupted by outliers. This paper addresses this problem with a novel, robust ML approach that is formulated as an optimization problem by coupling locally weighted least-squares support vector machines for regression (LWLS-SVMR) with one weight function. The weight is a function of residuals and allows for iteration within the proposed approach, significantly reducing the negative interference of outliers. A new efficient hybrid algorithm is developed to solve the optimization problem. The proposed approach is assessed and validated by comparison with relevant ML approaches on both one-dimensional simulated datasets corrupted by various outliers and multi-dimensional real-world engineering datasets, including datasets used for predicting the lateral strength of reinforced concrete (RC) columns, the fuel consumption of automobiles, the rising time of a servomechanism, and dielectric breakdown strength. Finally, the proposed method is applied to produce a data-driven solver for computational mechanics with a nonlinear material dataset corrupted by outliers. The results all show that the proposed method is robust against non-extreme and extreme outliers and improves the predictive performance necessary to solve various engineering problems.more » « less
-
ABSTRACT The heavily used protein–protein docking server ClusPro performs three computational steps as follows: (1) rigid body docking, (2) RMSD based clustering of the 1000 lowest energy structures, and (3) the removal of steric clashes by energy minimization. In response to challenges encountered in recent CAPRI targets, we added three new options to ClusPro. These are (1) accounting for small angle X‐ray scattering data in docking; (2) considering pairwise interaction data as restraints; and (3) enabling discrimination between biological and crystallographic dimers. In addition, we have developed an extremely fast docking algorithm based on 5D rotational manifold FFT, and an algorithm for docking flexible peptides that include known sequence motifs. We feel that these developments will further improve the utility of ClusPro. However, CAPRI emphasized several shortcomings of the current server, including the problem of selecting the right energy parameters among the five options provided, and the problem of selecting the best models among the 10 generated for each parameter set. In addition, results convinced us that further development is needed for docking homology models. Finally, we discuss the difficulties we have encountered when attempting to develop a refinement algorithm that would be computationally efficient enough for inclusion in a heavily used server. Proteins 2017; 85:435–444. © 2016 Wiley Periodicals, Inc.more » « less
-
Abstract We develop a convex‐optimization clustering algorithm for heterogeneous financial networks, in the presence of arbitrary or even adversarial outliers. In the stochastic block model with heterogeneity parameters, we penalize nodes whose degree exhibit unusual behavior beyond inlier heterogeneity. We prove that under mild conditions, this method achieves exact recovery of the underlying clusters. In absence of any assumption on outliers, they are shown not to hinder the clustering of the inliers. We test the performance of the algorithm on semi‐synthetic heterogenous networks reconstructed to match aggregate data on the Korean financial sector. Our method allows for recovery of sub‐sectors with significantly lower error rates compared to existing algorithms. For overlapping portfolio networks, we uncover a clustering structure supporting diversification effects in investment management.more » « less
-
Many graph problems can be solved using ordered parallel graph algorithms that achieve significant speedup over their unordered counterparts by reducing redundant work. This paper introduces a new priority-based extension to GraphIt, a domain-specific language for writing graph applications, to simplify writing high-performance parallel ordered graph algorithms. The extension enables vertices to be processed in a dynamic order while hiding low-level implementation details from the user. We extend the compiler with new program analyses, transformations, and code generation to produce fast implementations of ordered parallel graph algorithms. We also introduce bucket fusion, a new performance optimization that fuses together different rounds of ordered algorithms to reduce synchronization overhead, resulting in 1.2x--3x speedup over the fastest existing ordered algorithm implementations on road networks with large diameters. With the extension, GraphIt achieves up to 3x speedup on six ordered graph algorithms over state-of-the-art frameworks and hand-optimized implementations (Julienne, Galois, and GAPBS) that support ordered algorithms.more » « less
