Anastasia π
Anastasia πΒΆ
Title: Determining compounds with the highest electronic contribution to the dielectric constant
Description: The decrease of a transistorβs size allows doubling the number of transistors on a microchip every 18-24 months, which increases the computer performance without increasing power spendings. The current dielectric, SiO2, loses its insulating ability when too thin. Hence, there arises a problem of finding a new insulating material, particularly among compounds with a high dielectric constant, including a high electronic contribution to the dielectric constant. In this research project, we apply a machine learning approach to determine the electronic contribution to the dielectric constant for the compounds with undetermined dielectric properties. In particular, we use linear regression and random forests models with a variety of parameters to construct a list of the materials with the highest electronic contribution. As a next step of our work, we need to evaluate the possibility of mass production of the listed compounds, based on stability of the material, cost of production, and other industrial parameters.