Comparison of experimental measurements and machine learning predictions of dielectric constant of liquid crystals

Machine learning methods for liquid crystal research: phases, textures, defects and physical properties

“…One work stands apart from all the others mentioned in this section, because the authors analyze dielectric spectroscopy data to predict the real and imaginary parts of the dielectric constants of the phtalocyanine-doped nematic liquid crystal mixtures. 93 Two traditional regression algorithms ( k -NN and DT Regression) and five ensemble-based regression algorithms (Extreme Gradient Boosting, Random Forest, Extra Tree Regression, Voting and Bagging using k -Nearest Neighbor as a base learner) were tested for their predictive ability on a large experimental dataset of 1953 samples with three input parameters (frequency of an applied electric field, its voltage and dispersion rate) and two output parameters (the real and imaginary parts of the dielectric constant). It has been found that tree-based ensemble regression algorithms can best predict the dielectric properties of a composite liquid-crystalline material with an error rate of less than 5%.…”

Possibilities and limitations of convolutional neural network machine learning architectures in the characterisation of achiral orthogonal smectic liquid crystals

“…This has for example been demonstrated for the dielectric properties of a nematic LC through a comparison of the experimental and predicted values. 36 Another example is the prediction of elastic constants in relation to experimental and simulated curves. 37 Also melting temperatures have been shown to be predictable, 38 as has structural colour, i.e.…”

Machine learning classification of polar sub-phases in liquid crystal MHPOBC

“…30 Another issue, which is still in its infancy is the prediction of physical properties by machine learning, as demonstrated recently by a comparison of the experimental and predicted dielectric properties of a nematic liquid crystal. 31…”