Robust online active learning

Abstract

In many industrial applications, obtaining labeled observations is not straightforward as it often requires the intervention of human experts or the use of expensive testing equipment. In these circumstances, active learning can be highly beneficial in suggesting the most informative data points to be used when fitting a model. Reducing the number of observations needed for model development alleviates both the computational burden required for training and the operational expenses related to labeling. Online active learning, in particular, is useful in high‐volume production processes where the decision about the acquisition of the label for a data point needs to be taken within an extremely short time frame. However, despite the recent efforts to develop online active learning strategies, the behavior of these methods in the presence of outliers has not been thoroughly examined. In this work, we investigate the performance of online active linear regression in contaminated data streams. Our study shows that the currently available query strategies are prone to sample outliers, whose inclusion in the training set eventually degrades the predictive performance of the models. To address this issue, we propose a solution that bounds the search area of a conditional D‐optimal algorithm and uses a robust estimator. Our approach strikes a balance between exploring unseen regions of the input space and protecting against outliers. Through numerical simulations, we show that the proposed method is effective in improving the performance of online active learning in the presence of outliers, thus expanding the potential applications of this powerful tool.

A hybrid approach to soft sensor development for distillation-in-series plant under input data low variability

“…A total of 471 data points of the process were selected. The points from 1 to 353, corresponding to the period from 7 July 2022 to 25 October 2022, were included in the extended sample To demonstrate the effectiveness of the proposed approach (for predicting y on x i , i = 1, 6), we used a robust regression [31] and support vector machine (SVM) regression [32]. A cascade-forward neural network (CFN) and an ACE algorithm were used for nonparametric SS development.…”

The ENBIS‐22 quality and reliability engineering international special issue

“…Maculotti et al 3 objective is to compare different machine learning approaches with the aim of selecting the best one to predict the final quality of laser welds which allows to stick to non-destructive inspection methods as recommended by Industry 4.0. Cacciarelli et al 4 pose the original question of online active learning (learning algorithms that can actively query the user to obtain a label) in outliers-contaminated data streams; the method consists of both constraining the area of new labels and proposing a robust estimator. Misai et al 5 mix parametric, non-parametric, and machine learning inference methods to optimize a maintenance policy for a partially observed multi-component process.…”