Abstract

As the quantitative evaluation of dispersion of nanoparticles are crucial for the research of nanocomposites, including ceramic, metal, and polymer nanocomposites, we proposed two quantifiable tools for the evaluation of dispersion in the nanocomposites by using machine‐learning algorithms, that is the coefficient of variation of K‐nearest neighboring distances (CVKD value) and the information entropy derived from the probability density function. In total 230 different type of dispersion morphologies of nanocomposites were investigated by using K‐nearest neighboring algorithms, Gaussian mixture model, expectation maximization, and Kernel density estimation algorithms. The smaller values of CVKD or information entropy represented much more homogeneous dispersion of nanoparticles. It might be expected the information exhibited in the present study could assist in the rational design and fabrication of superior nanocomposites. POLYM. COMPOS., 40:1000–1005, 2019. © 2018 Society of Plastics Engineers

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call