Machine Learning for Technical Systems

A special issue of Mathematics (ISSN 2227-7390). This special issue belongs to the section "Network Science".

Deadline for manuscript submissions: closed (31 December 2022) | Viewed by 9550

Special Issue Editors

Center for Applied Data Science Gütersloh, Faculty of Engineering and Mathematics, FH Bielefeld - University of Applied Sciences, 33619 Bielefeld, Germany
Interests: machine learning; data science; neuroinformatics; high-performance computing; cognitive robotics
Center for Applied Data Science Gütersloh, Faculty of Engineering and Mathematics, FH Bielefeld - University of Applied Sciences, 33619 Bielefeld, Germany
Interests: machine learning; biometrics; data science; optimization

Special Issue Information

Dear Colleagues,

Functionality and features of technical systems, like assembly lines, robotic systems, motor vehicles, and imaging systems are becoming increasingly enhanced by data-driven machine learning algorithms. However, this application area imposes specific challenges. First of all, the safety of these systems needs to be guaranteed, requiring data-driven models that are interpretable and/or explainable after training, and that can be subject to thorough validation. Second, many technical systems generate continuous streams of data during operation, requiring advanced techniques for learning on data streams that can deal with concept drift and imbalanced data (e.g., error states may be rare but highly valuable for learning). Third, in such big data scenarios, the labeling of data is often expensive, raising the need to select the most representative or informative data points, i.e., to pursue “active learning”. Active learning is also highly valuable in small data scenarios for determining which data points to generate, whenever the generation of data points is expensive in terms of labor and financial costs. Fourth, most technical systems can be described by analytical models, and data-driven models obtained by machine learning may be used in a supplementary way for residual learning or as shortcut heuristics. Here, in this kind of hybrid modeling, the challenge lies in the integration of those analytical and data-driven models. Moreover, machine learning in technical systems often has to cope with real-time requirements and limited computing resources.

This Special Issue welcomes contributions that present novel algorithms or extensions of existing algorithms to meet these (and other) challenges in the application of machine learning to technical systems. Benchmarking of the proposed methods on data sets from real technical systems is most welcome.

Prof. Dr. Wolfram Schenck
Dr. Alaa Tharwat
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Mathematics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • Safety and validation in machine learning
  • Interpretability and explainability of data-driven models
  • Learning on data streams
  • Active learning
  • Hybrid modeling
  • Machine learning in real-time under resource constraints
  • Integration of domain-specific expert knowledge in machine learning models

Published Papers (4 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

32 pages, 4753 KiB  
Article
A Novel Low-Query-Budget Active Learner with Pseudo-Labels for Imbalanced Data
by Alaa Tharwat and Wolfram Schenck
Mathematics 2022, 10(7), 1068; https://doi.org/10.3390/math10071068 - 26 Mar 2022
Cited by 3 | Viewed by 1463
Abstract
Despite the availability of a large amount of free unlabeled data, collecting sufficient training data for supervised learning models is challenging due to the time and cost involved in the labeling process. The active learning technique we present here provides a solution by [...] Read more.
Despite the availability of a large amount of free unlabeled data, collecting sufficient training data for supervised learning models is challenging due to the time and cost involved in the labeling process. The active learning technique we present here provides a solution by querying a small but highly informative set of unlabeled data. It ensures high generalizability across space, improving classification performance with test data that we have never seen before. Most active learners query either the most informative or the most representative data to annotate them. These two criteria are combined in the proposed algorithm by using two phases: exploration and exploitation phases. The former aims to explore the instance space by visiting new regions at each iteration. The second phase attempts to select highly informative points in uncertain regions. Without any predefined knowledge, such as initial training data, these two phases improve the search strategy of the proposed algorithm so that it can explore the minority class space with imbalanced data using a small query budget. Further, some pseudo-labeled points geometrically located in trusted explored regions around the new labeled points are added to the training data, but with lower weights than the original labeled points. These pseudo-labeled points play several roles in our model, such as (i) increasing the size of the training data and (ii) decreasing the size of the version space by reducing the number of hypotheses that are consistent with the training data. Experiments on synthetic and real datasets with different imbalance ratios and dimensions show that the proposed algorithm has significant advantages over various well-known active learners. Full article
(This article belongs to the Special Issue Machine Learning for Technical Systems)
Show Figures

Figure 1

21 pages, 4688 KiB  
Article
Evaluation of sEMG Signal Features and Segmentation Parameters for Limb Movement Prediction Using a Feedforward Neural Network
by David Leserri, Nils Grimmelsmann, Malte Mechtenberg, Hanno Gerd Meyer and Axel Schneider
Mathematics 2022, 10(6), 932; https://doi.org/10.3390/math10060932 - 15 Mar 2022
Cited by 3 | Viewed by 2249
Abstract
Limb movement prediction based on surface electromyography (sEMG) for the control of wearable robots, such as active orthoses and exoskeletons, is a promising approach since it provides an intuitive control interface for the user. Further, sEMG signals contain early information about the onset [...] Read more.
Limb movement prediction based on surface electromyography (sEMG) for the control of wearable robots, such as active orthoses and exoskeletons, is a promising approach since it provides an intuitive control interface for the user. Further, sEMG signals contain early information about the onset and course of limb movements for feedback control. Recent studies have proposed machine learning-based modeling approaches for limb movement prediction using sEMG signals, which do not necessarily require domain knowledge of the underlying physiological system and its parameters. However, there is limited information on which features of the measured sEMG signals provide the best prediction accuracy of machine learning models trained with these data. In this work, the accuracy of elbow joint movement prediction based on sEMG data using a simple feedforward neural network after training with different single- and multi-feature sets and data segmentation parameters was compared. It was shown that certain combinations of time-domain and frequency-domain features, as well as segmentation parameters of sEMG data, improve the prediction accuracy of the neural network as compared to the use of a standard feature set from the literature. Full article
(This article belongs to the Special Issue Machine Learning for Technical Systems)
Show Figures

Figure 1

16 pages, 4786 KiB  
Article
Model-Assisted Labeling and Self-Training for Label Noise Reduction in the Detection of Stains on Images of Laundry
by Tamino Huxohl and Franz Kummert
Mathematics 2021, 9(19), 2498; https://doi.org/10.3390/math9192498 - 05 Oct 2021
Cited by 2 | Viewed by 1552
Abstract
In this work, the creation of a dataset labeled in a pixel-wise manner for the uncommon domain of stain detection on patterned laundry is described. The unique properties of images in this dataset—stains are small and sometimes occur in large amounts—led to the [...] Read more.
In this work, the creation of a dataset labeled in a pixel-wise manner for the uncommon domain of stain detection on patterned laundry is described. The unique properties of images in this dataset—stains are small and sometimes occur in large amounts—led to the creation of noisy labels. Indeed, the training of a fully convolutional neural network for salient object detection with this dataset revealed that the model predicts stains missed by human labelers. Thus, the reduction in label noise by adding overlooked regions with the help of the model’s predictions is examined in two different experiments. In the model-assisted labeling experiment, a simulation is ran where a human selects correct regions from the predictions. In the self-training experiment, regions of high certainty are automatically selected from the predictions. Re-training the model with the revised labels shows that model-assisted labeling leads to an average improvement in performance by 8.52%. In contrast, with self-training, the performance increase is generally lower (2.58% on average) and a decrease is even possible since regions of high certainty are often false positives. Full article
(This article belongs to the Special Issue Machine Learning for Technical Systems)
Show Figures

Figure 1

26 pages, 1134 KiB  
Article
Incremental DoE and Modeling Methodology with Gaussian Process Regression: An Industrially Applicable Approach to Incorporate Expert Knowledge
by Tim Voigt, Martin Kohlhase and Oliver Nelles
Mathematics 2021, 9(19), 2479; https://doi.org/10.3390/math9192479 - 04 Oct 2021
Cited by 2 | Viewed by 2732
Abstract
The use of data-based models is a favorable way to optimize existing industrial processes. Estimation of these models requires data with sufficient information content. However, data from regular process operation are typically limited to single operating points, so industrially applicable design of experiments [...] Read more.
The use of data-based models is a favorable way to optimize existing industrial processes. Estimation of these models requires data with sufficient information content. However, data from regular process operation are typically limited to single operating points, so industrially applicable design of experiments (DoE) methods are needed. This paper presents a stepwise DoE and modeling methodology, using Gaussian process regression that incorporates expert knowledge. This expert knowledge regarding an appropriate operating point and the importance of various process inputs is exploited in both the model construction and the experimental design. An incremental modeling scheme is used in which a model is additively extended by another submodel in a stepwise fashion, each estimated on a suitable experimental design. Starting with the most important process input for the first submodel, the number of considered inputs is incremented in each step. The strengths and weaknesses of the methodology are investigated, using synthetic data in different scenarios. The results show that a high overall model quality is reached, especially for processes with few interactions between the inputs and low noise levels. Furthermore, advantages in the interpretability and applicability for industrial processes are discussed and demonstrated, using a real industrial use case as an example. Full article
(This article belongs to the Special Issue Machine Learning for Technical Systems)
Show Figures

Figure 1

Back to TopTop