Lowering Barriers to AI Adoption with AutoML and Interpretability

Overview

Building good machine learning models is difficult and time consuming, and few engineers and scientists have the necessary experience. Automated Machine Learning (AutoML) simplifies that process to a few steps, identifying the best model and optimizing its hyperparameters in a single step, thus making machine learning accessible to any engineer. We will also demonstrate various interpretability methods available in MATLAB that overcome the black box nature of machine learning, lowering the bar to adoption of machine learning in industries that cannot tolerate black box models, including Finance and Medical applications. Finally, we explain how incremental learning makes models improve over time and adopt to changing conditions.

Highlights

• Learn the three steps to obtain an optimized predictive model from raw signal or image data
• Demonstrate various interpretability methods to explain model predictions
• Apply incremental learning to adapt models to changes in the environment

About the Presenter

Bernhard Suhm is the product manager for Machine Learning at MathWorks. He works closely with customer facing and development teams to address customer needs and market trends in our machine learning related products, primarily the Statistics and Machine Learning toolbox. Prior to joining MathWorks Bernhard applied analytics to optimizing the delivery of customer service in call centers, after specializing in speech user interfaces in his PhD from Carnegie Mellon and Karlsruhe University (Germany).