Via Science and the math that drives Big Data
April 21, 2014![default image](https://nucleusresearch.com/wp-content/themes/nucleus/assets/images/blog-inside-featured.png)
I had an opportunity to meet with the folks at Via Science recently and came away impressed with their technology. The problem with Big Data is, well, that it’s big. You need processing power to crunch the data and more than a bit of knowledge about the data to create models. For instance, rain and umbrella sales are linked, but umbrella sales do not make it rain, just as sales of 4×8 sheets of plywood do not attract hurricanes. With large data sets finding the links becomes increasingly more difficult. That’s where Via Science comes in. Using Bayesian Networks, a combination of graphical models and probability, the folks at Via Science can crunch large data sets to find meaningful correlations. I’m simplifying the problem but the end result is uncovering the factors that influence the outcome and generating models that can be used to predict or optimize an outcome. Right now Via Science’s solution is used by organizations doing heavy lifting with Big Data but it’s clear the technology will eventually find favor right down to the desktop. A nice prediction of the future of analytics.