Copyright © 2018 DataScience.US All Rights Reserved.
Self-service Data Discovery and Interactive Predictive Analytics
What’s so important about the convergence of self-service data exploration with interactive predictive analytics? Why does this matter to your organization?
Think of all the self-service things you use in a day. Gas pumps. ATMs. Online apps for stock trading, banking and shopping. The idea behind them is added convenience for consumers and reduced costs for providers. People choose what they want, when they want – without involving others in their minute-to minute decisions.
The same goes for self-service data exploration and easy-to-use analytics. Increasingly, people want to explore data and search for answers – when they want, where they want and how they want to – without asking IT for data or reports. We’re seeing it even in everyday data applications like the Fitbit or Apple Watch™. People are hungry for information they can find, see and use on their own.
[Tweet “”The data fabric is the next middleware.” –Todd Papaioannou”]
So whether it’s combining caloric intake with exercise output for that magic fitness number, digging into why weekly sales are trending downwards from the office or drilling deep into a report from a mobile device at an airport, giving people the opportunity to explore data on their own empowers them to make better decisions (e.g., skip the doughnut at the airport).
But it’s not just about learning what happened and why. What about the future? Adding interactive predictive analytics to your organizational mix enables decision makers and data consumers to go beyond the past to discovering (for themselves) insights about the future.
Combining interactive predictive analytics with self-service data discovery also enables statisticians and data scientists to explore data relationships before they dive into model building and refinement. They can quickly and iteratively try different data and techniques to find the model that produces the best answer. When satisfied with the model, they can easily collaborate with business analysts to explore the modeling results from the same environment. Once identified, the champion model can be put into production by embedding it into operational applications and scoring new data to produce faster insights.
It’s becoming clear that three factors are important for organizations that want to provide real value from data to everyone who needs it:
- Interactivity and ease of use for users of all skill levels.
- Powerful predictive analytics and consumable results.
- An architecture that can handle big data and the need for speed.
Reference: Gartner. “Survey Analysis: Customers Rate Their Advanced Analytics Platforms