Vitech Hacklahom Challenge
Vitech's Hacklahoma 2018 Challenge
About our Application
Using a Neural Network with Back-propogation (written in TFLearn), we were able to predict insurance quotes and plans that consumers will purchase. We developed Python scripts to pull data from a REST web service and compiled a local copy for the neural network to train on. Currently, with 45,000 data points on consumer data, we were about to predict the price of the different tiers of insurance as well as the plan the consumer purchased with an 82% accuracy. This could have been improved by improving the network structure further if given enough time and more data points.
Log in or sign up for Devpost to join the conversation.