Artificial Intelligence (AI) techniques have already been penetrated widely and deeply into our surroundings. Particularly, in 2016, Google Deepmind's AI model AlphaGo and Lee Sedol match attracted great attention worldwide because it is the confrontation between humans and machines.
The autonomous driving technology can be an example to real life. The era of self-driving cars has come true. Currently, we have a level of technological prowess that rarely requires human assistance while driving.
For other cases, the AI techniques have been applied to various real-life fields, such as smart diagnostic kits, radiology, quiz shows. Speech recognition technology, which has recently been in the spotlight, is also a field among AI that is developing faster with deep learning algorithms.
Machine learning is one of the most useful techniques in materials science fields for developing new materials.
Recently, Materials Square updated the CGCNN module. It applied the Crystal Graph Convolutional Neural Network (CGCNN) model, the representative prediction model in the materials science field. CGCNN is a direct design model, which uses material structures as an input to predict material properties as an output. In other words, we can predict various basic properties (formation energy, band gap, bulk/shear moduli, Poisson's ratio) from structural information.
Materials Square's graphical user interface (GUI) allows researchers to be even unfamiliar with simulation to accumulate data required for machine learning easily. Applying the obtained high-quality materials science data to the CGCNN model would lead to new results with higher accuracy.
Start materials development using the CGCNN module!
Do you need more Information?