Predicting molecular ordering in a binary liquid crystal using machine learning

Inokuchi, Takuya; Okamoto, Ryosuke; Arai, Noriyoshi

doi:10.1080/02678292.2019.1656293

Cited by 20 publications

(13 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Aer a number of the above-mentioned studies predicting the properties of monodisperse molecular systems, a similar attempt was made for polydisperse systems. 45 (Fig. 6b) A…”

Section: Prediction Of Liquid Crystal Phases Based On Molecularlevel ...mentioning

confidence: 96%

“…After a number of the above-mentioned studies predicting the properties of monodisperse molecular systems, a similar attempt was made for polydisperse systems. 45 (Fig. 6b) A dissipative particle dynamics simulation method, developed specifically for soft matter and complex liquids, 52 was used to analyse self-assemble structures and phase transitions in the binary LC system.…”

Section: Machine Learning For Liquid Crystalsmentioning

confidence: 99%

See 1 more Smart Citation

Machine learning for soft and liquid molecular materials

et al. 2023

View full text Add to dashboard Cite

show abstract

“…Aer a number of the above-mentioned studies predicting the properties of monodisperse molecular systems, a similar attempt was made for polydisperse systems. 45 (Fig. 6b) A…”

Section: Prediction Of Liquid Crystal Phases Based On Molecularlevel ...mentioning

confidence: 96%

Section: Machine Learning For Liquid Crystalsmentioning

confidence: 99%

Machine learning for soft and liquid molecular materials

et al. 2023

View full text Add to dashboard Cite

show abstract

“…On a broader dataset of LCs, consisting of 243 five-ring aromatic compounds with varying terminal chains, linkage groups, and substituents on central and outer rings, a Multi-Automated Regression Splines model was able to achieve an R-squared value of 0.900 using 2-dimensional and 3-dimensional descriptors, performing better than the Support Vector Machine (Antanasijevic et al, 2016). From 15 input features describing the structure of an array of LC beads, a random forest regression model was able to find the phase transition temperature with a coefficient of determination of 0.943, and from those 15-input feature plus the phase transition temperature, was able to predict the order parameter with a coefficient of determination of 0.894 (Inokuchi et al, 2019).…”

Section: Using Qsprs For ML Of Physical Properties Of Lcsmentioning

confidence: 98%

The Applications of Machine Learning in the Study of Liquid Crystals: A Review

Kalinin

Abercrombie

2023

J Stud Res

View full text Add to dashboard Cite

Discovered in the 19th century, liquid crystals have only grown more prominent in the last thirty to forty years as a result of electric fields being able to influence their nematic directors. Since these nematic directors can polarize light, liquid crystals have found extensive uses in the field of optoelectronics, including being in many of the devices (such as computers, phones, and tablets) that we have today. With the continued desire to improve these optoelectronic devices comes the desire to improve the liquid crystal systems that help to compose them. The growth of machine learning technologies has been near the forefront of these efforts to improve the efficiency and effectiveness of the development of these systems. In an effort to recognize this prominence, this article presents a review of machine learning’s presence and use in the study of liquid crystals, as well as a comparison of machine learning techniques and more traditional experimental methods.

show abstract

“…In contrast to hard particle systems, dynamic modeling of dissipative particles was used in ref. 45 to represent two different types of liquid crystal molecules in binary mixtures, namely molecules consisting of 5 and 10 beads. Among the five regression methods (RF, linear regression, ridge regression, elastic net regression, and support vector regression) that predict the order parameter and phase transition temperature for self-assembling structures modeled by coarse-grained molecular modeling, the RF method showed the best prediction results with the determination coefficient R 2 = 0.894 for the order parameters and R 2 = 0.943 for the transition temperature.…”

Section: Main Textmentioning

confidence: 99%