Quick Insights Into Big Data With PowerBI & Machine Learning

UPDATE: I see on LinkedIn that Gartner Business Intelligence and Analytics Magic Quadrant 2017 has been released recently showing Microsoft continues to be the leader in terms of vision and ability to execute. I do encourage you to read the full report here but the one specific take away is here:

Microsoft is recognized for the constant and fast development of Microsoft Power BI. This is the 10th consecutive year that Microsoft has been positioned as a leader.

powerbiFollowing on from my last post that referenced how the Cortana Intelligence Suite was powering Sticky Notes into the 21st Century through the use of machine learning, I’ve just seen a relatively new feature in PowerBI that is doing the same thing. It’s called Quick Insights and you can read all about it on the following link:

Quick Insights With PowerBI

The basic overview is that after you’ve published your data set from PowerBI Desktop to PowerBi in the Azure cloud, you can either start to manually build some reports for your data and analyse it with questions you may have OR you can use Quick Insights. As per the website:

The Quick Insights feature is built on a growing set of advanced analytical algorithms developed in conjunction with Microsoft Research that we’ll continue to use to allow more people to find insights in their data in new and intuitive ways.

To see this in action, watch the following video showing some examples:

The reality is I think the ability of advanced analytical algorithms to find trends or outliers “hidden” in data will probably exceed the abilities of amateur “data scientists” who are doing their best to pick these out. I recall a post on LinkedIn from Dr Joe Sweeney where he talked about the real value in big data being in the algorithms and the companies that can develop those the fastest/best will be in a strong position.

It will be interesting to see how smart these Quick Insights end up being when end users, particularly schools who typically don’t have full time data scientists working for them, start to use them to interrogate their data.

Talking of examining data, this is a good example from Microsoft’s Ray Fleming showing how the natural language Q&A feature can drill down deep into your data and auto-magically format it for you:

This example is not using machine learning, but instead leveraging well named fields in the data’s table structures to quickly locate and visualise data.

I am always keen to discuss what I've written and hear your ideas so leave a reply here...

%d bloggers like this: