At the risk of being accused of only using Amazon Web Services, here is a look at modeling using Microsoft Azure Machine Learning Studio along with the R programming language. It is chock-full of data munging, modeling, and delivery functions!
If you want to follow along, sign up at AzureML for a free account.
Part 1 - Simple Demo with the Adult Census Income Binary Classification Dataset
Sign into your account at https://studio.azureml.net and follow the following steps:
Click EXPERIMENTS in the left horizontal menu bar:
Click + NEW sign at bottom left of screen to start a new experiment
Choose Blank experiment:
Time to drag-and-drop modeling modules onto the workspace
Select Saved Datasets then Samples and drag Adult Census Income Binary Classification dataset onto your workspace:
Your workspace should look like:
You can right click on the data set to visualize the data.
Select Data Transformation then Sample and Split and drag Split Datat onto your workspace:
And connect both rectangles together:
Select Machine Learning then Train and drag Train Model onto your workspace:
And connect both rectangles as such:
You'll notice the red exclamation point in the last rectangle - this is because you need to tell it what feature is your outcome variable. So click on the red exclamation point and in the right menu pane choose 'income' as outcome:
Select Machine Learning then Initialize Model then Classification
and drag Two-Class Logistic Regression onto your workspace:
And connect it to the top left of Train Model:
Select Machine Learning then Score
and drag Score Model onto your workspace:
Connect Train Model to top left Score Model and connect Split Data to top right of Score Model. Fun connecting all this, right?:
Select Machine Learning then Evaluate
and drag Evaluate Model onto your workspace:
Connect bottom Score Model to top left Evaluate Model:
Finally, hit RUN button at bottom middle of Azure ML Studio page:
Right click on Evaluate Model and click Evaluation Results - Visualize
Pretty simple and easy to use, right? Play around with the Threshold slider - great to understand the flexibility and cost of the AUC score.