Regression Analysis of Superconductivity Data

Abstract

The dataset used in this project is a superconductivity dataset introduced by Hamidieh (2018) and made available through the UCI Machine Learning Repository. It contains in formation on 21,263 superconductors, each represented by 81 derived features that summarize important elemental properties such as atomic radius, valence, thermal conductivity, electron affinity, and atomic mass. These features were calculated as statistical aggregates (mean, standard deviation, range, etc.) based on the chemical composition of each material. The target variable is the critical temperature (Tc), the point at which the material becomes superconducting. The data is multivariate with real-valued attributes, and the main task associated with it is regression. The purpose of this project is to develop predictive models capable of estimating the critical temperature of superconductors from these derived material properties.