Skip to main content



Highly accurate predictive model gives educators insights to increase STEM student retention


NLP Logix, on the heels of finishing 13th out of 821 teams from across the globe who built data models to accurately predict which students would drop out of a massive online course in the KDD Cup 2015, has announced that it is offering this model to educators involved in STEM initiatives across the United States.

“We spent countless hours training the data to produce a model which was able to, with very high accuracy, predict which student would drop out of a specific on-line course.” said Robert Marsh, Co-Founder and Chief Information Officer, NLP Logix. “Now we have taken that same technology and have made it available to be applied to just about any Learning Management System (LMS) to give educators actionable insights as to which students may have a higher probability of dropping out of a specific course.”

NLP Logix, which is one of the fastest growing predictive analytics and machine learning companies in the country, is leveraging the success they have had not only in the KDD Cup 2015 competition, but also other education-based contracts they have been performing upon over the past few years. One of these contracts is with the University of Florida to measure the effects of education for the Florida Poison Information Center Network and the results were recently presented to the Florida Legislature.

“We really have a great infrastructure to be able to take large education-based data sets and develop highly accurate predictive models,” said Austin Seymour, Business Intelligence Engineer, NLP Logix. “Not only can we build the models quickly using our LogixStudio machine learning tool kit, but we are able to take those results and build visualizations using Tableau Software and deliver them in a flash.”

NLP Logix has begun work for a major university and is in advanced discussions with a number of others across the Southeast.

Leave a Reply