Metric Learning with Missing Data
Amaury Lendasse
University of Houston, USA
Abstract
This talk will discuss the problem of metric learning, including variable selection and variable ranking in the presence of missing data (also known as incomplete data). Problems with incomplete data and missing values are common and important in real-world machine learning scenarios, but yet often underrepresented in the research field. Particularly data related to healthcare tends to feature missing values that must be handled properly, and ignoring incomplete samples is not an acceptable solution.
In this talk, we will focus on classification problems with a large number of variables and with every sample being incomplete. We have a triple goal: to deal with the missing data, to achieve high accuracy, and to bring some interpretability and explainability to the classification problems. A quick review of metric learning will also be presented.
Speaker
Amaury Lendasse was born in Tournai, Belgium. He received an M.S. degree in Mechanical Engineering from the Université catholique de Louvain (Belgium) in 1996, an M.S. in Control in 1997, and a Ph.D. in Applied Mathematics (Engineering) in 2003 from the same university. In 2003, he was a postdoctoral researcher in the Computational Neurodynamics Lab at the University of Memphis. From 2004 to 2014, he was a Professor pro tempore and Senior Researcher in Computer Science at the Aalto University in Finland. During that period, he created and led the Environmental and Industrial Machine Learning (EIML) at Aalto. From 2014 to 2019, he was an Associate Professor at The University of Iowa (USA). He is now a Full Professor, a faculty senator, and a Department Chair at the University of Houston (USA) in the Department of Information and Logistics Technology (ILT) in the College of Engineering. He is a visiting Professor at Arcada University of Applied Sciences in Finland. He was the Chairman of the annual ESTSP conference (European Symposium on Time Series Prediction) and a member of the editorial board and program committee of several journals and conferences on machine learning. He is the author or co-author of more than 300 scientific papers in international journals, books, or communications to conferences with reviewing committees. His research includes Data Science, Machine Learning, Big Data, time series prediction, variable selection, and missing data. According to Google Scholar, he has ~9300 citations and an h-index of 44. At the International Conference on Extreme Learning Machines in Yantai (China) in October 2017, he received the Pioneer Award for his contributions to Machine Learning and Extreme Learning Machines.