Identifying Medicare Fraud Using Automated Machine Learning

Sep 14, 2016 · Cambridge, United States of America

6:00 - 6:30 PM - Networking

6:30 - 8 PM - Talk and Q&A

Title & Abstract:

Identifying Medicare Fraud Using Automated Machine Learning

Over the past year, Boston startup DataRobot has been quietly building the industry’s best-automated machine learning platform, and they are finally ready to unveil the software to the Boston Data Mining meetup community. 

Tom de Godoy, co-founder of Data Robot  will demonstrate how  DataRobot's automated machine learning platform delivers the best predictive model for your data. He will take the audience through pre-processing and feature engineering all the way through testing and tuning, using the best open source algorithms from R, Python, Spark, TensorFlow, H2O, and more.

During the session, Tom de Godoy will model data, extracted from the Office of Inspector General and the Center for Medicare and Medicaid to show how DataRobot would identify doctors most likely to commit Medicare fraud.

After picking the most accurate machine learning model for identifying fraud, the speaker will leverage Data Robot's data visualization techniques, which provide deeper insights into the model output. Finally, DataRobot's  novel technique will be adopted to summarize the model’s evidence against each high-risk physician.

Speaker Bio:

Tom de Godoy is co-founder and CTO of DataRobot, and has over 10 years of Data Science and engineering experience. Previously, he was Senior Director of Research and Modeling at Travelers Insurance where he managed a team of Data Scientists on applications in pricing, claims and customer behavior for various insurance products. Tom has been ranked as high as 20th in the world on the data science competition platform, which boasts more than 500,000 registered Data Scientists.

Event organizers

Are you organizing Identifying Medicare Fraud Using Automated Machine Learning?

Claim the event and start manage its content.

I am the organizer

based on 0 reviews