An adaptive fused sampling approach of high-accuracy data in the presence of low-accuracy data

Abstract

In several applications, a large amount of Low-Accuracy (LA) data can be acquired at a small cost. However, in many situations, such LA data is not sufficient for generating a higidelity model of a system. To adjust and improve the model constructed by LA data, a small sample of High-Accuracy (HA) data, which is expensive to obtain, is usually fused with the LA data. Unfortunately, current techniques assume that the HA data is already collected and concentrate on fusion strategies, without providing guidelines on how to sample the HA data. This work addresses the problem of collecting HA data adaptively and sequentially so when it is integrated with the LA data a more accurate surrogate model is achieved. For this purpose, we propose an approach that takes advantage of the information provided by LA data as well as the previously selected HA data points and computes an improvement criterion over a design space to choose the next HA data point. The performance of the proposed method is evaluated, using both simulation and case studies. The results show the benefits of the proposed method in generating an accurate surrogate model when compared to three other benchmarks.

Keywords:

• Data fusion
• adaptive sampling
• expected improvement criterion

Additional information

Notes on contributors

Mostafa Reisi Gahrooei

Mostafa Reisi Gahrooei is a Ph.D. student in the Department of Industrial and Systems Engineering (ISyE) at Georgia Institute of Technology. Mostafa Reisi Gahrooei completed a master's degree in computational science and engineering at Georgia Tech, and received double M.Sc. degrees in transportation engineering and applied mathematics both from Southern Illinois University Edwardsville. His research focuses on modeling and monitoring of systems with heterogeneous, high-dimensional data. Mr. Reisi Gahrooei is a member of the Institute for Operations Research and the Management Sciences (INFORMS) and the Institute of Industrial and Systems Engineers (IISE).

Kamaran Paynabar

Kamran Paynabar received a B.Sc. degree from the Iran University of Science and Technology, Tehran, Iran, in 2002, and an M.Sc. degree in industrial engineering from Azad University, Tehran, Iran, in 2004, an M.A. degree in statistics and the Ph.D. degree in industrial and operations engineering in 2012 from the University of Michigan, Ann Arbor, MI, USA .Currently, he is an assistant professor at the H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, GA, USA. His research interests include data fusion for multi-stream waveform signals and functional data, engineering-driven statistical modeling, sensor selection in distributed sensing networks, probabilistic graphical models, and statistical learning with applications in manufacturing and healthcare systems. Dr. Paynabar is a Member of the Institute of Industrial and Systems Engineers (IISE) and The Institute for Operations Research and the Management Sciences (INFORMS).

Massimo Pacella

Massimo Pacella is an assistant professor in the Department of Engineering for Innovation at the University of Salento, Lecce Italy, where he teaches statistical process control and design of experiments. He received an M.Sc. in computerengineering from the University of Lecce Italy, and a Ph.D. degree in manufacturing and production systems from the Polytechnic of Milan Italy, in 1998 and 2003, respectively. In 2009, he was awarded a Fulbright Fellowship. In 2013 and 2017, he received the Italian National Scientific Qualification for the role of associate professor in manufacturing technology and systems. His major research interests are in functional data processing, profile monitoring, design of experiments, manufacturing process control and coordinate metrology, including the development of artificial intelligence techniques and methods of applied statistics. He is member of the Italian Association for Manufacturing Technology (AITeM).

Bianca Maria Colosimo

Bianca Maria Colosimo is a professor in the Department of Mechanical Engineering at the Politecnico di Milano. Her research interests include statistical process modeling and monitoring for big data in advanced manufacturing. She is a senior member of the American Society for Quality and Editor-elect of Journal of Quality Technology. Her email address is [email protected].