Capture Blog

How To Cheat at Enrollment Predictions, Part 4: Predicting Enrollments Using Prospects

Welcome to the grand finale of “How To Cheat at Enrollment Predictions,” a series that sprang out of a lively conversation among Capture Higher Ed’s data scientists and members of the company’s intrepid — and let’s face it vocal — sales team.

The question: How do Capture’s competitors make their predictive models look better than they really are? In other words, how do they cheat at prediction?

We narrowed it down to four major ways: 1) using leakers, 2) overfitting data; 3) predicting close to the target date and …

Today we look at the fourth way: Predicting enrollments using prospects.

From a large pool, only a very small number of prospects will enroll in your school. But have no fear. This actually makes your prediction task easier if you approach it … like some of our competitors do, including the industry’s largest player.

Let’s say .2 percent of all your prospects will enroll in your school. If you predicted a 0 percent likelihood of enrollment for everyone, you’d be 99.8 percent accurate. That prediction isn’t very useful, but it sure looks good.

You certainly wouldn’t want to act like Capture and hurt your accuracy statistics by predicting the next major stage in the process — application. While that would be a heck of a lot more useful in the real world, it might not look as good on paper. And if you want to cheat at predictions, do whatever it takes to make your model look good on paper.

Accurately predicting the future? That’s a problem for your future self.

With Envision, Capture has developed education’s first ensemble predictive model, unique to the enrollment management industry because it combines various machine-learning algorithms to build ensemble predictive models for application, enrollment and financial aid.

Unlike current predictive models in higher education, which are built upon historical data and not easily iterated upon, our patent-pending models use machine learning and multiple algorithms to continuously look for patterns and make more accurate predictions every time new data becomes available.

See for yourself what we are talking about. Take the Envision Challenge and let us run your data for free. We’re eager to show you what we can do.

By John Foster, Data Analyst, Capture Higher Ed

Share this!Share on FacebookTweet about this on TwitterShare on Google+Share on LinkedIn
Please wait...

Subscribe to our weekly blog digest!

You will get fresh weekly content, and we will never sell your email.