Upside Briefing: Rulex Analytics
Rulex provides a new generation of advanced analytics tools. We offer Upside's first impressions from a business briefing.
- By Fern Halper
- April 18, 2016
Company: Rulex Analytics
Company Location: Boston, MA, USA; Genova, Italy
TDWI Product Categories: Advanced Analytics, Machine Learning, Predictive Analytics
Company Vision: Providing a new generation of advanced analytics tools for proactive decision makers.
Briefing Notes: I had the opportunity to speak with Andrea Ridi (CEO) and Tim Negris (VP Marketing) from Rulex Analytics. Rulex provides what it terms "cognitive machine learning" software to help end users from across the organization -- data scientists as well as business analysts -- find rules that exist in big data. According to Rulex, in cognitive machine learning the software examines historical data and finds conditional logic patterns -- if-then rules -- in the data. Rulex learns by inferring if-then rules from the data and then selects the most relevant rule set. These decision rules are continuously re-evaluated against new events to ensure persistent accuracy. Rules can be refreshed when they go stale.
In the demo, the user selected his/her data set(s) in Rulex Data Worker -- the automated data discovery piece of Rulex. Users can blend and profile and transform the data and specify the outcome variable of interest (e.g., a categorical variable such as leave or stay or a number). This is all done via a GUI interface that is part of Data Worker.
Rulex Data Worker also identifies the most relevant combinations of input values in the data. According to the company, Data Worker can be used with traditional analytics tools, such as SAS STAT, for logistic regression, and it can be used with conventional machine-learning algorithms such as neural networks and support vector machines. Rulex Data Modeler takes the data identified by Data Worker and produces the if-then statements using an algorithm called a Logic Learning Machine. These rules can be input to the Rulex Forecaster scoring engine or parsed and processed by application frameworks as well as end point devices.
Currently, Rulex can manage 1 billion cells on a 32 gigabyte machine.
First Impressions: Rulex Analytics is a relatively new entrant into the field of what I'm calling democratized data science/predictive analytics. This is the idea that you don't have to be a statistician or a mathematician to be able to build more advanced ML and predictive models. Other players in this space include BeyondCore and Watson Analytics. In fact, many analytics vendors are trying to make their advanced analytics easier to use.
I found a few things interesting about Rulex. First, Rulex produces the simplest set of rules for the data. Additionally, the Rulex rules are readable in plain English -- the users don't need to know how to read the output of a machine learning algorithm like a decision tree. Now, a decision tree is also a machine-learning algorithm that can also evaluate historical data to produce if-then rules. What's different here is that in a decision tree the rules don't overlap. That means you can end up with a large set of specific rules.
Second, I appreciated the fact that the software was designed to be deployed; i.e., to take action on the results and that included embedding the scoring engine in endpoint devices.
Fern Halper, Ph.D., is well known in the analytics community, having published hundreds of articles, research reports, speeches, webinars, and more on data mining and information technology over the past 20 years. Halper is also co-author of several “Dummies” books on cloud computing, hybrid cloud, and big data. She is the director of TDWI Research for advanced analytics, focusing on predictive analytics, social media analysis, text analytics, cloud computing, and “big data” analytics approaches. She has been a partner at industry analyst firm Hurwitz & Associates and a lead analyst for Bell Labs. Her Ph.D. is from Texas A&M University. You can reach her at firstname.lastname@example.org, on Twitter @fhalper, and on LinkedIn at linkedin.com/in/fbhalper.