jaikrishnap / machine-learning-for-prediction-of-bcfas Goto Github PK

Once they are available, we should link the dataset and our publication at the appropriate places in the README file.

Some people might not want to use conda, so we should probably also include instructions on how to setup a virtual environment with virtualenv and include a requirements.txt file for pip.

It would be nice if the script could handle missing values in some of the input features concerning the size of the aggregate. This would also allow joint processing of csv files where some rows specify, e.g., the number of primary particles and others the mobility diameter.

There should be a section that explains how to train you own model on some dataset and use it with the prediction script. Maybe we can just link to the other repo that contains the experiments.