Easy to Use and Continuous Learning Automated Machine Learning

Abstract

Machine Learning is becoming a tool that is used to solve many problems from finance, to health to education. Organizations look forward to having data scientists that will be building, maintaining and updating the machine learning solutions. Yet, a machine learning solution is much more difficult to maintain than any other piece of software, because it keeps evolving with data. Only with the help of technologies, such as explainability, continuous learning, learning from human and AutoML and with the involvement of business experts, it is going to be viable to have as many machine learning models as needed. The machine learning models will be supporting decisions of human experts and allowing them to make faster and more accurate decisions. In this talk, I will introduce these concepts and also give a demo of the AutoML platform that we developed at tazi.ai.

Bio

Zehra Cataltepe is the CEO and Co-Founder of Tazi, one of four worldwide Gartner “Cool Vendor in AI Core Technologies” (May 2019). At Tazi, her aim is to help transform Machine Learning into a technology that can be used easily and useful for everyone. After she got her Ph.D. from Caltech in Computer Science, Zehra worked at Bell Labs and Siemens Corporate Research. At Istanbul Technical University, she has been teaching graduate level Machine Learning and Pattern Recognition courses each year and has supervised Ph.D.and M.Sc. theses. She has 18 patents and over 90 publications on theory of machine learning and its applications in finance, health, energy and transportation.

Talk Time and Place

MacQuarrie Hall Room 225 @ 1:30 on March 10, 2020