Bin Jiang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Bin Jiang has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
Anthony Rhodes
Fariborz Maseeh Department of Mathematics and Statistics Portland State University Portland OR USA
2 shared publicationsJenny Jiang
Goizueta Business School Emory University Atlanta GA USA
2 shared publications- Peng Jiang
Department of Chemical Engineering, University of Florida, Gainesville, Florida 32611, United States
2 shared publications Abigail M. Fenton
Department of Chemical Engineering, University of Florida, Gainesville, FL 32611, USA
1 shared publication- Anthony D. Rhodes
Portland State University Portland State University
1 shared publication - Christopher Kim
Mechanical Engineering, Northwestern University , Evanston, IL 60208 , USA
1 shared publication
Latest Publications
- 21 October 2024
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 32 (1)
Graph Regularized Sparse L2,1 Semi‐Nonnegative Matrix Factorization for Data Reduction
Non‐negative Matrix Factorization (NMF) is an effective algorithm for multivariate data analysis, including applications to feature selection, pattern recognition, and computer vision. Its variant, Semi‐Nonnegative Matrix Factorization (SNF), extends the ability of NMF to render parts‐based data representations to include mixed‐sign data. Graph Regularized SNF builds upon this paradigm by adding a graph regularization term to preserve the local geometrical structure of the data space. Despite their successes, SNF‐related algorithms to date still suffer from instability caused by the Frobenius norm due to the effects of outliers and noise. In this paper, we present a new SNF algorithm that utilizes the noise‐insensitive norm. We provide monotonic convergence analysis of the SNF algorithm. In addition, we conduct numerical experiments on three benchmark mixed‐sign datasets as well as several randomized mixed‐sign matrices to demonstrate the performance superiority of SNF over conventional SNF algorithms under the influence of Gaussian noise at different levels.
- 21 October 2024
- Published by Wiley in Numerical Linear Algebra with Applications
- Vol. 32 (1)
Graph Regularized Sparse L2,1 Semi‐Nonnegative Matrix Factorization for Data Reduction
Non‐negative Matrix Factorization (NMF) is an effective algorithm for multivariate data analysis, including applications to feature selection, pattern recognition, and computer vision. Its variant, Semi‐Nonnegative Matrix Factorization (SNF), extends the ability of NMF to render parts‐based data representations to include mixed‐sign data. Graph Regularized SNF builds upon this paradigm by adding a graph regularization term to preserve the local geometrical structure of the data space. Despite their successes, SNF‐related algorithms to date still suffer from instability caused by the Frobenius norm due to the effects of outliers and noise. In this paper, we present a new SNF algorithm that utilizes the noise‐insensitive norm. We provide monotonic convergence analysis of the SNF algorithm. In addition, we conduct numerical experiments on three benchmark mixed‐sign datasets as well as several randomized mixed‐sign matrices to demonstrate the performance superiority of SNF over conventional SNF algorithms under the influence of Gaussian noise at different levels.
- Open AccessGraph Regularized Sparse $L_{2,1}$ Semi-Nonnegative Matrix Factorization for Data ReductionNon-negative Matrix Factorization (NMF) is an effective algorithm for multivariate data analysis, including applications to feature selection, pattern recognition, and computer vision. Its variant, Semi-Nonnegative Matrix Factorization (SNF), extends the ability of NMF to render parts-based data representations to include mixed-sign data. Graph Regularized SNF builds upon this paradigm by adding a graph regularization term to preserve the local geometrical structure of the data space. Despite their successes, SNF-related algorithms to date still suffer from instability caused by the Frobenius norm due to the effects of outliers and noise. In this paper, we present a new $L_{2,1}$ SNF algorithm that utilizes the noise-insensitive $L_{2,1}$ norm. We provide monotonic convergence analysis of the $L_{2,1}$ SNF algorithm. In addition, we conduct numerical experiments on three benchmark mixed-sign datasets as well as several randomized mixed-sign matrices to demonstrate the performance superiority of $L_{2,1}$ SNF over conventional SNF algorithms under the influence of Gaussian noise at different levels.Open AccessGraph Regularized Sparse $L_{2,1}$ Semi-Nonnegative Matrix Factorization for Data ReductionNon-negative Matrix Factorization (NMF) is an effective algorithm for multivariate data analysis, including applications to feature selection, pattern recognition, and computer vision. Its variant, Semi-Nonnegative Matrix Factorization (SNF), extends the ability of NMF to render parts-based data representations to include mixed-sign data. Graph Regularized SNF builds upon this paradigm by adding a graph regularization term to preserve the local geometrical structure of the data space. Despite their successes, SNF-related algorithms to date still suffer from instability caused by the Frobenius norm due to the effects of outliers and noise. In this paper, we present a new $L_{2,1}$ SNF algorithm that utilizes the noise-insensitive $L_{2,1}$ norm. We provide monotonic convergence analysis of the $L_{2,1}$ SNF algorithm. In addition, we conduct numerical experiments on three benchmark mixed-sign datasets as well as several randomized mixed-sign matrices to demonstrate the performance superiority of $L_{2,1}$ SNF over conventional SNF algorithms under the influence of Gaussian noise at different levels.
- Open Access
Regularized L21-Based Semi-NonNegative Matrix Factorization
We present a general-purpose data compression algorithm, Regularized L21 Semi-NonNegative Matrix Factorization (L21 SNF). L21 SNF provides robust, parts-based compression applicable to mixed-sign data for which high fidelity, individualdata point reconstruction is paramount. We derive a rigorous proof of convergenceof our algorithm. Through experiments, we show the use-case advantages presentedby L21 SNF, including application to the compression of highly overdeterminedsystems encountered broadly across many general machine learning processes.
Open AccessRegularized L21-Based Semi-NonNegative Matrix Factorization
We present a general-purpose data compression algorithm, Regularized L21 Semi-NonNegative Matrix Factorization (L21 SNF). L21 SNF provides robust, parts-based compression applicable to mixed-sign data for which high fidelity, individualdata point reconstruction is paramount. We derive a rigorous proof of convergenceof our algorithm. Through experiments, we show the use-case advantages presentedby L21 SNF, including application to the compression of highly overdeterminedsystems encountered broadly across many general machine learning processes.
- Open Access
- 23 January 2020
- Published by Wiley in Numerical Methods for Partial Differential Equations
- Vol. 36 (5) , 1129-1144
An enhanced finite difference time domain method for two dimensional Maxwell's equations
An enhanced finite‐difference time‐domain (FDTD) algorithm is built to solve the transverse electric two‐dimensional Maxwell's equations with inhomogeneous dielectric media where the electric fields are discontinuous across the dielectric interface. The new algorithm is derived based upon the integral version of the Maxwell's equations as well as the relationship between the electric fields across the interface. To resolve the instability issue of Yee's scheme (staircasing) caused by discontinuous permittivity across the interface, our algorithm revises the permittivities and makes some corrections to the scheme for the cells around the interface. It is also an improvement over the contour‐path effective permittivity algorithm by including some extra terms in the formulas. The scheme is validated in solving the scattering of a dielectric cylinder with exact solution from Mie theory and is then compared with the above contour‐path method, the usual staircasing and the volume‐average method. The numerical results demonstrate that the new algorithm has achieved significant improvement in accuracy over other methods. Furthermore, the algorithm has a simple structure and can be merged into current FDTD software packages easily. The C++ source code for this paper is provided as supporting information for public access.
- 23 January 2020
- Published by Wiley in Numerical Methods for Partial Differential Equations
- Vol. 36 (5) , 1129-1144
Open AccessAn enhanced finite difference time domain method for two dimensional Maxwell's equations
An enhanced finite‐difference time‐domain (FDTD) algorithm is built to solve the transverse electric two‐dimensional Maxwell's equations with inhomogeneous dielectric media where the electric fields are discontinuous across the dielectric interface. The new algorithm is derived based upon the integral version of the Maxwell's equations as well as the relationship between the electric fields across the interface. To resolve the instability issue of Yee's scheme (staircasing) caused by discontinuous permittivity across the interface, our algorithm revises the permittivities and makes some corrections to the scheme for the cells around the interface. It is also an improvement over the contour‐path effective permittivity algorithm by including some extra terms in the formulas. The scheme is validated in solving the scattering of a dielectric cylinder with exact solution from Mie theory and is then compared with the above contour‐path method, the usual staircasing and the volume‐average method. The numerical results demonstrate that the new algorithm has achieved significant improvement in accuracy over other methods. Furthermore, the algorithm has a simple structure and can be merged into current FDTD software packages easily. The C++ source code for this paper is provided as supporting information for public access.
- Open Access
- 9 December 2019
- Published by IOP Publishing in Materials Research Express
- Vol. 7 (1) , 015603
Simulation to fabrication—understanding the effect of NiAuCu alloy catalysts for controlled growth of graphene at reduced temperature
It is a significant challenge to grow large-scale, high quality, monolayer graphene at low temperature for the applications in industry, especially for the complementary metal oxide semiconductor fabrication process. To overcome this difficulty, we simulated the decomposition of acetylene (C2H2) on (100) surfaces of primarily nickel (Ni) catalysts with small mol fractions of gold (Au) and copper (Cu), using a 4 × 4 × 4 periodic supercell model. Based on the calculation of the reaction energies to decompose the C-H or C≡C bonds on different catalyst surfaces, a differential energy is proposed to clearly scale the decomposition difficulties such that larger differential energy leads to easier control of the monolayer growth. It is observed that on the NiAuCu alloy surface with a mol fraction 0.0313 of both Au and Cu, the differential energy of the C-H bonds and the C≡C bond are both positive, showing an obvious modulation effect on the decomposition of C2H2 and the catalytic activites. The simulation result is consistent with the growth of uniform monolayer graphene on silicon dioxide substrate at 500°C by plasma enhanced chemical vapor deposition with C2H2 precursor and Ni alloy catalysts with 1 wt% Au and 1 wt% Cu.
- 9 December 2019
- Published by IOP Publishing in Materials Research Express
- Vol. 7 (1) , 015603
Open AccessSimulation to fabrication—understanding the effect of NiAuCu alloy catalysts for controlled growth of graphene at reduced temperature
It is a significant challenge to grow large-scale, high quality, monolayer graphene at low temperature for the applications in industry, especially for the complementary metal oxide semiconductor fabrication process. To overcome this difficulty, we simulated the decomposition of acetylene (C2H2) on (100) surfaces of primarily nickel (Ni) catalysts with small mol fractions of gold (Au) and copper (Cu), using a 4 × 4 × 4 periodic supercell model. Based on the calculation of the reaction energies to decompose the C-H or C≡C bonds on different catalyst surfaces, a differential energy is proposed to clearly scale the decomposition difficulties such that larger differential energy leads to easier control of the monolayer growth. It is observed that on the NiAuCu alloy surface with a mol fraction 0.0313 of both Au and Cu, the differential energy of the C-H bonds and the C≡C bond are both positive, showing an obvious modulation effect on the decomposition of C2H2 and the catalytic activites. The simulation result is consistent with the growth of uniform monolayer graphene on silicon dioxide substrate at 500°C by plasma enhanced chemical vapor deposition with C2H2 precursor and Ni alloy catalysts with 1 wt% Au and 1 wt% Cu.
- 1 August 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 7, 1-3
Single Bandpass Microwave Photonic Filter based on Gain-switched Distributed-feedback Laser
In this paper, a single bandpass microwave photonic filter (MPF) based on a gain-switched distributed feedback (DFB) laser is proposed and experimentally demonstrated. Self-phase modulation (SPM) effect is explored in the optical domain to enhance the performance of the proposed MPF. A single passband is achieved for the proposed MPF, which can be tuned from 1 GHz to 39 GHz. A theoretical model describing the principle of the proposed MPF is derived, which is confirmed by experimental results.
- 1 August 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 7, 1-3
Single Bandpass Microwave Photonic Filter based on Gain-switched Distributed-feedback Laser
In this paper, a single bandpass microwave photonic filter (MPF) based on a gain-switched distributed feedback (DFB) laser is proposed and experimentally demonstrated. Self-phase modulation (SPM) effect is explored in the optical domain to enhance the performance of the proposed MPF. A single passband is achieved for the proposed MPF, which can be tuned from 1 GHz to 39 GHz. A theoretical model describing the principle of the proposed MPF is derived, which is confirmed by experimental results.
- 1 January 2017
- Published by Elsevier in Thin Solid Films
- Vol. 621, 156-164
Scalable parallel self-assembly of nanoparticle anti-reflection coatings
- 1 January 2017
- Published by Elsevier in Thin Solid Films
- Vol. 621, 156-164
Scalable parallel self-assembly of nanoparticle anti-reflection coatings
- 19 June 2007
- Published by Wiley in Numerical Methods for Partial Differential Equations
- Vol. 24 (2) , 476-503
A priori error analysis of the mortar finite element method for variational inequalities
- 19 June 2007
- Published by Wiley in Numerical Methods for Partial Differential Equations
- Vol. 24 (2) , 476-503
A priori error analysis of the mortar finite element method for variational inequalities
- 1 December 2006
- Published by Elsevier in Computer Methods in Applied Mechanics and Engineering
- Vol. 196 (1-3) , 371-378
Convergence analysis of P1 finite element method for free boundary problems on non-overlapping subdomains
- 1 December 2006
- Published by Elsevier in Computer Methods in Applied Mechanics and Engineering
- Vol. 196 (1-3) , 371-378
Convergence analysis of P1 finite element method for free boundary problems on non-overlapping subdomains
- 20 April 2005
- Published by Wiley in Numerical Methods for Partial Differential Equations
- Vol. 22 (1) , 1-17
A nonoverlapping domain decomposition method for variational inequalities derived from free boundary problems
- 20 April 2005
- Published by Wiley in Numerical Methods for Partial Differential Equations
- Vol. 22 (1) , 1-17
A nonoverlapping domain decomposition method for variational inequalities derived from free boundary problems