Machine Learning Enabled Capacitance Prediction for Carbon-Based Supercapacitors

07 May 2018, Version 2
This content is a preprint and has not undergone peer review at the time of posting.


Carbon is the most widely used electrode for the supercapacitors. To predict the capacitance of carbon-based supercapacitors, this work applies three machine learning (ML) methods, including linear regression, Lasso and artificial neural network. For training the ML process, we extracted data from hundreds of published papers. Moreover, five variables were selected to figure out their impact on capacitance, including specific surface area, calculated pore size, ID/IG ratio, N-doping level and voltage window. By evaluated with the real data, all of three methods achieve acceptable prediction results, and ANN exhibits the best performance. More importantly, this work shows the potential of ML in material science and advanced applications.


machine learning
artificial neural network
performance prediction

Supplementary materials

Machine Learning ChemRxiv20180504.2


Comments are not moderated before they are posted, but they can be removed by the site moderators if they are found to be in contravention of our Commenting Policy [opens in a new tab] - please read this policy before you post. Comments should be used for scholarly discussion of the content in question. You can find more information about how to use the commenting feature here [opens in a new tab] .
This site is protected by reCAPTCHA and the Google Privacy Policy [opens in a new tab] and Terms of Service [opens in a new tab] apply.