IEEE Austin EMBS & Computer and ComSoc & SP Joint Austin chapters invites you to a special presentation on Machine Learning and Healthcare
Date: November 29, 2018
Time: 6-8 Pm
Location: AT&T Labs, 9505 Arboretum, Austin
Topic: “BEST OF TWO WORLDS: PERFORMANCE AND EXPLAINABILITY IN MACHINE LEARNING FOR HEALTHCARE APPLICATIONS”
PLEASE RSVP – (This is required to access the AT&T building)
Machine learning can help health professionals and patients make better, faster, and more cost-effective decisions. Still, in many situations simply having a machine learning model “spit out” a prediction without a clear explanation of the reasons behind it won’t produce the desired behavior. Historically, solving this problem required adopting predictive models that can be succinctly summarized, such as small decision trees. This often meant sacrificing the accuracy that could be achieved with more complex models like neural networks or ensembles, whose inner workings are too difficult to comprehend.
This session will cover:
- Scenarios where explainability become critical for an effective use of machine learning models in healthcare settings.
- Creative approaches to alleviate the inherent tension between performance and explainability and avoid unsatisfactory trade-offs in both directions.
- The ROI and how-to of explainability for complex machine learning models.
About the speaker:
Adriana Beal has 25 years tech industry experience, ranging from IBM and HP to early-stage start-ups. Recent roles: Director, Product Management @ Gravitant, Data Analytics Consultant @ BP3 Global. Core expertise: machine learning, data analytics, and cloud computing. Educational background: B.Sc. Electronic Engineering, MBA in Strategic Management of Information, Data Analytics Certification.
For any details, please email F.Behmann@IEEE.org
Chair, ComSoc/SP and EMBS/CS Austin Chapters