This week, over 400 engineers, scientists, and mathematicians met at the BioScience Research Collaborative building at Rice University to discuss the advances, promises, and challenges of machine learning. The day was a remarkable display of the cross-disciplinary impact that this growing field has had in finance, energy, law, and healthcare.
What was the format?
ML@Rice was jointly hosted by Rice University’s Ken Kennedy Institute and the Departments of Electrical & Computer Engineering, Computer Science, and Computational & Applied Mathematics. A mixture of academics and industry professionals held talks ranging from theory to commercial application, and the local ML community showed off some of the latest advances in the field with an afternoon poster session. Jointly hosted by Rice University’s Ken Kennedy Institute and the Departments of Electrical & Computer Engineering, Computer Science, and Computational & Applied Mathematics,
What our co-founder had to say
As part of the workshop, MIC’s co-founder and CTO, Dr. Craig Rusin, was invited to speak about the use of machine learning in healthcare. Each speaker presented their own experience using ML in their work. Craig spoke about using ML to save lives. He explained the difficulty in obtaining the data that he needed for his medical research, and how Sickbay allowed him to gather this data and then use it to train an algorithm that predicts cardiac arrest 1-2 hours in advance.
As the day proceeded, speakers repeatedly demonstrated the breadth of machine learning’s applications in their fields. For example, in healthcare, Genevera Allen described the new statistical models that her research group created to calculate the relationships between various genomic data sets. A few hours later, Dr. Hardeep Singh provided the audience with a wish list of advances that incorporate social science into learning from the EMR to provide better diagnoses and follow-up.
Despite the broad range of the industries addressed by the workshop, common ground was reached surprisingly often. During a late afternoon panel of industry experts, we learned that restricted data access limits progress not only in healthcare, but in oil and gas. While privacy concerns and trade secrets are rightfully held as important reasons for maintaining data silos, advances in secure and large-scale machine learning by researchers like Anshumali Shrivastava point the way towards better algorithms trained on larger, richer data sets shared securely from multiple sources.
MIC’s future plans
Of course, MIC has collected one of the largest, richest physiologic data sets in the world. And we are just getting started. We are proud to be a part of the ever-increasing ML community in Space City, and are already thinking about our plans for the next ML@Rice.
If you’d like to learn more about how we use machine learning to save lives, Contact MIC today!