Best models for Binary Classification RRS feed

  • Question

  • Hi everyone!

    I am a college student enrolled in a class called "Introduction to Business Analytics", and come here to see if you could help me with a group task that involves using Azure ML. The case is as follows:

    We have a .xml file that shows data from 10k cases in which is stated if a company has been sanctioned for any violation of data protection in the last 5 years. We had to build a Azure ML model in order to predict if the company is going to be fined or not. This model uses 80% of the data for training and 20% for testing.

    We've tried various models of binary classification and obtained a max AUC score of 0.793 with the Two-Class Neural Network Model. Can we improve our AUC score by changing some of the parameters? Or.. are there better predictive models? We still don't understand how the score fluctuates depending on the model/parameters.

    I guess this is very basic, but we're thrilled to learn more about AzureML.

    A response that could shed a little bit of light would be amazing!!

    Thank you in advance

    Friday, November 8, 2019 4:36 PM

All replies