Software

Software:

NP-ODE: Neural Process Aided Ordinary Differential Equations for Uncertainty Quantification of Finite Element Analysis: [Link]

  • This package presents a physics-informed surrogate model, named Neural Process Aided Ordinary Differential Equation.

  • More details can be found in the paper [Wang, Y., Wang, K., Cai, W. and Yue, X., 2020. NP-ODE: Neural Process Aided Ordinary Differential Equations for Uncertainty Quantification of Finite Element Analysis. arXiv preprint arXiv:2012.06914.]

StressNet: A Deep Learning method to Predict Stress with Fracture Propagation in Brittle Materials.

  • The code and data are available upon request.

  • More details can be found in the paper [Wang*, Y., Oyen, D., Guo, W., Mehta, A., Scott, C.B., Panda, N., Fernandez-Godino, M.G., Srinivasan, G., Yue, X., (2021) "StressNet - Deep learning to predict stress with fracture propagation in brittle materials", NPJ Materials Degradation, 5, 6]

CPAC-CNN: Tensor decomposition to Approximately Compress Convolutional Layers in Deep Learning: [Link]

  • This package presents how to use Tensor decomposition to compress convolutional kernels in deep learning.

  • Compared with the original convolutional layer, the proposed CPAC-Conv layer can reduce the number of parameters without decaying prediction performance.

  • More details can be found in the paper preprint [Wang, Y., Guo, W., Yue, X., (2020+) "CPAC-Conv: CP-decomposition to Approximately Compress Convolutional Layers in Deep Learning", arXiv:2005.13746]

Penalized Mixed-Effects Decomposition Package: [Link] [Code Ocean]

  • This package can be used to decompose multi-channel functional data into multiple components.

  • More details can be found in the paper [Yue, X., et al., 2018. “A Wavelet-based Penalized Mixed-Effects Decomposition for Multichannel Profile Monitoring based on In-line Raman Spectroscopy”, IEEE Transactions on Automation Science and Engineering, 15(3), pp.1258-1271]

Tensor Mixed-Effects Decomposition Package

  • This package can be requested by email currently. We are preparing the standardized package and will make it public soon.

Datasets:

Raman Spectra Dataset for the Single-Wall Carbon Nanotubes Buckypaper: [Link]

  • A Single-Walled Carbon Nanotubes (SWCNT) buckypaper sample was fabricated and measured with the Raman spectroscopy. The acquisition time for static Raman spectroscopy varies from 0.01 to 1 s (including 0.01s, 0.05s, 0.1s, 0.5s, 1s ) for the SWCNT buckypaper.

  • In all measurements, near infrared laser with a wavelength of 785 nm and a power of 150 mW was used to eliminate the effect of ambient light. Low-magnification lens was used to achieve larger focus tolerance.

  • More details can be found in paper [Yue, X., et al., 2017. Generalized Wavelet Shrinkage of Inline Raman Spectroscopy for Quality Monitoring of Continuous Manufacturing of Carbon Nanotube Buckypaper. IEEE Transactions on Automation Science and Engineering, 14(1), pp.196-207.]