A confidence booster from HAL24K’s data scientists

Here at HAL24K we reckon we’re pretty good at predicting things, and we hope that our customers would agree. However as good as our models are, we know they can never be perfect. Because of this we’re always looking for ways of quantifying our model’s performance and providing this information to our customers.

In the latest post on our TechBlog, Sam Blake, one of our senior data scientists, shows how we go about generating confidence intervals for our existing deep learning models - no retraining required!

Read the blog here. And keep an eye out as we plan to post more content on tech insights and open sourcing in the months ahead.