Azure Machine Learning Studio Office Add-In

I am currently in the process of learning more about Azure Machine Learning Studio . Within there you can create and train predictive models. As I am going through some of the examples I came across a great example of an Office Add-In.

I am creating a simple weather predicting example based on data I downloaded from an API service. More on that later. From my model I created a sample Web Service with the click of a button. This exposes an End point for me to send data to be “predicted on”.

a2

When you create a Web Service from your trained model you can access a test screen through the application.

a1

 

From there I spotted a link to an Excel App and I figure I would see what was going on there. Turns out it is an Office Add-In… !!

When the Excel sheet opened I was presented with the option to test the web service

a3

Once you load your test data into the sheet you can then use the Add-In to select the data to test – and designate where the answer will be added back to the sheet.

a4

And you push Predict !!!

Looking at the application using the F12 tool we can see the hosted Office Add-In location and also see the prediction in action.

a7

a5

The web service is called and the response is a JSON string with all the answers and the scored confidence in the answer.

The Add-In then inserts those answers back into the Excel spreadsheet for you

a6

As you can see, my predicting needs some work but the Add-In and sample web service worked like a charm !!!

Poking around the site I can see it is built using Knockout.js and some jQuery – very cool 🙂

a8

Once I have properly figured out how this all Machine Learning lark works I am going to create some blog posts on the predictive side of this – but I wanted to share how cool this out of the box for free Add-In is

GREAT USE CASE !!

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s