CO2_H2O_prediction

Introduction

We are aiming to improve the prediction error of carbon dioxide and water concentration in silicate melts.

Data Description

The algorithm was trained on experiments that have the concentrations of various compounds documented, such as SiO2, TiO2, etc. There is an example input template in the Data folder.

Methodology

We are buidling a weighted linear regression based model in Python with statsmodels module.

Conclusion

The algorithm is able to predict carbon dioxide concentration with 5% error in terms of ppm, and 10% for water concentration.