Data Prediction

by Andrew H
I read an article by, Alex Williams called “Prior Knowledge: A Predictive Database for Developers.” In his article Alex talks about the company Prior Knowledge that debuted their predictive database application called Veritable on September 11th in San Francisco. The Prior Knowledge team has spent years learning the art of statistics so that they could build their knowledge into this software. As a result they were able to develop a software that allows developers to build applications that can determine what things are. One of Alex’s examples of what the company uses it for is working with retailers to determine their customers purchasing patterns. He also states, that the service “magically fills in values that may be missing.” The end result will allow data to be analyzed that other wise would not be able to.

The way the software analyzes casual relationships and is able to withdraw useful information is great. This can open doors for many businesses and take a lot of work off of the data analysts as most of their data will have already be predicted for them allowing them to make better use of their time analyzing what the software cannot. If this software continues to grow and developers are able to build applications with Veritable API then companies will have a totally new way of looking at what otherwise could be useless data.

I thought this article was very interesting as I have not read anything about predictive data before. I know that we have not talked about predicting data values in class; however I think this software is very important and that we should talk about it because if it takes off it will be an imporant part of database applications in the future.

Williams, A. (2012, September 11). Prior Knowledge: A Predictive Database for Developers. Retrieved from techcrunch:

3 thoughts on “Data Prediction”

  1. I thought that this article was really interesting in that it related to my own article. The fact that someone is developing software to determine customer’s purchasing patterns probably must have prior information in order to determine this. My article was about how companies are gathering data based on the search histories, interests, and behaviors of its users. I thought that the information gathered would be necessary for the software analysis process of the program.

  2. It’s a really interesting article. I haven’t read anything about predictive data before either. I did some research on this, for SQL data, from a table, PredictiveDB allows you to predict the NULL values of a column. But, PredictiveDB associates an error risk to its predictions. Nice post~

  3. Good choice of article. Predictive modeling/Predictive analytics can be applied to various industries/fields, such as estimating demand, or predicting churn (when customers are about to move out of product/service) in telecom; estimating customer lifetime value and credit risk, targeting customers in financial services; improving health care and government services; organizing air traffic in transportation; and determining the customized next best offer in eCommerce (e.g. Amazon, Netflix).

Comments are closed.