Sensors

Research

Jump to: Review

15 pages, 3260 KiB

Open AccessArticle

Machine Learning Models for the Automatic Detection of Exercise Thresholds in Cardiopulmonary Exercising Tests: From Regression to Generation to Explanation

by Andrea Zignoli

Sensors 2023, 23(2), 826; https://doi.org/10.3390/s23020826 - 11 Jan 2023

Cited by 4 | Viewed by 2352

Abstract

The cardiopulmonary exercise test (CPET) constitutes a gold standard for the assessment of an individual’s cardiovascular fitness. A trend is emerging for the development of new machine-learning techniques applied to the automatic process of CPET data. Some of these focus on the precise [...] Read more.

The cardiopulmonary exercise test (CPET) constitutes a gold standard for the assessment of an individual’s cardiovascular fitness. A trend is emerging for the development of new machine-learning techniques applied to the automatic process of CPET data. Some of these focus on the precise task of detecting the exercise thresholds, which represent important physiological parameters. Three are the major challenges tackled by this contribution: (A) regression (i.e., the process of correctly identifying the exercise intensity domains and their crossing points); (B) generation (i.e., the process of artificially creating a CPET data file ex-novo); and (C) explanation (i.e., proving an interpretable explanation about the output of the machine learning model). The following methods were used for each challenge: (A) a convolutional neural network adapted for multi-variable time series; (B) a conditional generative adversarial neural network; and (C) visual explanations and calculations of model decisions have been conducted using cooperative game theory (Shapley’s values). The results for the regression, generation, and explanatory techniques for AI-assisted CPET interpretation are presented here in a unique framework for the first time: (A) machine learning techniques reported an expert-level accuracy in the classification of exercise intensity domains; (B) experts are not able to substantially differentiate between a real vs an artificially generated CPET; and (C) Shapley’s values can provide an explanation about the choices of the algorithms in terms of ventilatory variables. With the aim to increase their technology-readiness level, all the models discussed in this contribution have been incorporated into a free-to-use Python package called pyoxynet (ver. 12.1). This contribution should therefore be of interest to major players operating in the CPET device market and engineering. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

18 pages, 4400 KiB

Open AccessArticle

Application of Machine Learning Methods for an Analysis of E-Nose Multidimensional Signals in Wastewater Treatment

by Magdalena Piłat-Rożek, Ewa Łazuka, Dariusz Majerek, Bartosz Szeląg, Sylwia Duda-Saternus and Grzegorz Łagód

Sensors 2023, 23(1), 487; https://doi.org/10.3390/s23010487 - 02 Jan 2023

Cited by 13 | Viewed by 2043

Abstract

The work represents a successful attempt to combine a gas sensors array with instrumentation (hardware), and machine learning methods as the basis for creating numerical codes (software), together constituting an electronic nose, to correct the classification of the various stages of the wastewater [...] Read more.

The work represents a successful attempt to combine a gas sensors array with instrumentation (hardware), and machine learning methods as the basis for creating numerical codes (software), together constituting an electronic nose, to correct the classification of the various stages of the wastewater treatment process. To evaluate the multidimensional measurement derived from the gas sensors array, dimensionality reduction was performed using the t-SNE method, which (unlike the commonly used PCA method) preserves the local structure of the data by minimizing the Kullback-Leibler divergence between the two distributions with respect to the location of points on the map. The k-median method was used to evaluate the discretization potential of the collected multidimensional data. It showed that observations from different stages of the wastewater treatment process have varying chemical fingerprints. In the final stage of data analysis, a supervised machine learning method, in the form of a random forest, was used to classify observations based on the measurements from the sensors array. The quality of the resulting model was assessed based on several measures commonly used in classification tasks. All the measures used confirmed that the classification model perfectly assigned classes to the observations from the test set, which also confirmed the absence of model overfitting. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

16 pages, 4031 KiB

Open AccessArticle

Mutual-Aided INS/Vision Navigation System Analysis and Optimization Using Sequential Filtering with State Recalculation

by Ayham Shahoud, Dmitriy Shashev and Stanislav Shidlovskiy

Sensors 2023, 23(1), 79; https://doi.org/10.3390/s23010079 - 22 Dec 2022

Viewed by 1963

Abstract

This paper presents the implementation of a mutual-aided navigation system for an aerial vehicle. Employing all available sensors in navigation is effective at maintaining continuous and optimal results. The images offer a lot of information about the surrounding environment, but image processing is [...] Read more.

This paper presents the implementation of a mutual-aided navigation system for an aerial vehicle. Employing all available sensors in navigation is effective at maintaining continuous and optimal results. The images offer a lot of information about the surrounding environment, but image processing is time-consuming and causes timing problems. While traditional fusion algorithms tend to reduce the delay errors or ignore them, this research depends on state estimation recalculation during the delay time and on sequential filtering. To reduce the image matching time, the map is processed offline, then key point clusters are stored to avoid feature recalculation online. The sensors’ information is used to bound the search space for the matched features on the map, then they are reprojected on the captured images to exclude the unuseful part from processing. The suggested mutual-aided form compensates for the inertial system drift, which enhances the system’s accuracy and independence. The system was tested using data collected from a real flight using a DJI drone. The measurements from an inertial measurement unit (IMU), camera, barometer, and magnetometer were fused using a sequential Kalman Filter. The final results prove the efficiency of the suggested system to navigate with high independency, with an RMS position error of less than 3.5 m. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

16 pages, 5845 KiB

Open AccessArticle

Path Generator with Unpaired Samples Employing Generative Adversarial Networks

by Javier Maldonado-Romo, Alberto Maldonado-Romo and Mario Aldape-Pérez

Sensors 2022, 22(23), 9411; https://doi.org/10.3390/s22239411 - 02 Dec 2022

Cited by 3 | Viewed by 1219

Abstract

Interactive technologies such as augmented reality have grown in popularity, but specialized sensors and high computer power must be used to perceive and analyze the environment in order to obtain an immersive experience in real time. However, these kinds of implementations have high [...] Read more.

Interactive technologies such as augmented reality have grown in popularity, but specialized sensors and high computer power must be used to perceive and analyze the environment in order to obtain an immersive experience in real time. However, these kinds of implementations have high costs. On the other hand, machine learning has helped create alternative solutions for reducing costs, but it is limited to particular solutions because the creation of datasets is complicated. Due to this problem, this work suggests an alternate strategy for dealing with limited information: unpaired samples from known and unknown surroundings are used to generate a path on embedded devices, such as smartphones, in real time. This strategy creates a path that avoids virtual elements through physical objects. The authors suggest an architecture for creating a path using imperfect knowledge. Additionally, an augmented reality experience is used to describe the generated path, and some users tested the proposal to evaluate the performance. Finally, the primary contribution is the approximation of a path produced from a known environment by using an unpaired dataset. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

14 pages, 678 KiB

Open AccessArticle

Optimal Greedy Control in Reinforcement Learning

by Alexander Gorobtsov, Oleg Sychev, Yulia Orlova, Evgeniy Smirnov, Olga Grigoreva, Alexander Bochkin and Marina Andreeva

Sensors 2022, 22(22), 8920; https://doi.org/10.3390/s22228920 - 18 Nov 2022

Viewed by 1169

Abstract

We consider the problem of dimensionality reduction of state space in the variational approach to the optimal control problem, in particular, in the reinforcement learning method. The control problem is described by differential algebraic equations consisting of nonlinear differential equations and algebraic constraint [...] Read more.

We consider the problem of dimensionality reduction of state space in the variational approach to the optimal control problem, in particular, in the reinforcement learning method. The control problem is described by differential algebraic equations consisting of nonlinear differential equations and algebraic constraint equations interconnected with Lagrange multipliers. The proposed method is based on changing the Lagrange multipliers of one subset based on the Lagrange multipliers of another subset. We present examples of the application of the proposed method in robotics and vibration isolation in transport vehicles. The method is implemented in FRUND—a multibody system dynamics software package. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

27 pages, 6099 KiB

Open AccessArticle

Design of a Forest Fire Early Alert System through a Deep 3D-CNN Structure and a WRF-CNN Bias Correction

by Alejandro Casallas, Camila Jiménez-Saenz, Victor Torres, Miguel Quirama-Aguilar, Augusto Lizcano, Ellie Anne Lopez-Barrera, Camilo Ferro, Nathalia Celis and Ricardo Arenas

Sensors 2022, 22(22), 8790; https://doi.org/10.3390/s22228790 - 14 Nov 2022

Cited by 9 | Viewed by 3389

Abstract

Throughout the years, wildfires have negatively impacted ecological systems and urban areas. Hence, reinforcing territorial risk management strategies against wildfires is essential. In this study, we built an early alert system (EAS) with two different Machine Learning (ML) techniques to calculate the meteorological [...] Read more.

Throughout the years, wildfires have negatively impacted ecological systems and urban areas. Hence, reinforcing territorial risk management strategies against wildfires is essential. In this study, we built an early alert system (EAS) with two different Machine Learning (ML) techniques to calculate the meteorological conditions of two Colombian areas: (i) A 3D convolutional neural net capable of learning from satellite data and (ii) a convolutional network to bias-correct the Weather Research and Forecasting (WRF) model output. The results were used to quantify the daily Fire Weather Index and were coupled with the outcomes from a land cover analysis conducted through a Naïve-Bayes classifier to estimate the probability of wildfire occurrence. These results, combined with an assessment of global vulnerability in both locations, allow the construction of daily risk maps in both areas. On the other hand, a set of short-term preventive and corrective measures were suggested to public authorities to implement, after an early alert prediction of a possible future wildfire. Finally, Soil Management Practices are proposed to tackle the medium- and long-term causes of wildfire development, with the aim of reducing vulnerability and promoting soil protection. In conclusion, this paper creates an EAS for wildfires, based on novel ML techniques and risk maps. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

22 pages, 1184 KiB

Open AccessArticle

Online Multivariate Anomaly Detection and Localization for High-Dimensional Settings

by Mahsa Mozaffari, Keval Doshi and Yasin Yilmaz

Sensors 2022, 22(21), 8264; https://doi.org/10.3390/s22218264 - 28 Oct 2022

Cited by 3 | Viewed by 1945

Abstract

This paper considers the real-time detection of abrupt and persistent anomalies in high-dimensional data streams. The goal is to detect anomalies quickly and accurately so that the appropriate countermeasures could be taken in time before the system possibly gets harmed. We propose a [...] Read more.

This paper considers the real-time detection of abrupt and persistent anomalies in high-dimensional data streams. The goal is to detect anomalies quickly and accurately so that the appropriate countermeasures could be taken in time before the system possibly gets harmed. We propose a sequential and multivariate anomaly detection method that scales well to high-dimensional datasets. The proposed method follows a nonparametric, i.e., data-driven, and semi-supervised approach, i.e., trains only on nominal data. Thus, it is applicable to a wide range of applications and data types. Thanks to its multivariate nature, it can quickly and accurately detect challenging anomalies, such as changes in the correlation structure. Its asymptotic optimality and computational complexity are comprehensively analyzed. In conjunction with the detection method, an effective technique for localizing the anomalous data dimensions is also proposed. The practical use of proposed algorithms are demonstrated using synthetic and real data, and in variety of applications including seizure detection, DDoS attack detection, and video surveillance. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

17 pages, 639 KiB

Open AccessArticle

Real-Time Detection and Classification of Power Quality Disturbances

by Mahsa Mozaffari, Keval Doshi and Yasin Yilmaz

Sensors 2022, 22(20), 7958; https://doi.org/10.3390/s22207958 - 19 Oct 2022

Cited by 13 | Viewed by 2058

Abstract

This paper considers the problem of real-time detection and classification of power quality disturbances in power delivery systems. We propose a sequential and multivariate disturbance detection method (aiming for quick and accurate detection). Our proposed detector follows a non-parametric and supervised approach, i.e., [...] Read more.

This paper considers the problem of real-time detection and classification of power quality disturbances in power delivery systems. We propose a sequential and multivariate disturbance detection method (aiming for quick and accurate detection). Our proposed detector follows a non-parametric and supervised approach, i.e., it learns nominal and anomalous patterns from training data involving clean and disturbance signals. The multivariate nature of the method enables joint processing of data from multiple meters, facilitating quicker detection as a result of the cooperative analysis. We further extend our supervised sequential detection method to a multi-hypothesis setting, which aims to classify the disturbance events as quickly and accurately as possible in a real-time manner. The multi-hypothesis method requires a training dataset per hypothesis, i.e., per each disturbance type as well as the ’no disturbance’ case. The proposed classification method is demonstrated to quickly and accurately detect and classify power disturbances. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

23 pages, 3319 KiB

Open AccessArticle

Intelligent Traffic Monitoring through Heterogeneous and Autonomous Networks Dedicated to Traffic Automation

by Eduard Zadobrischi

Sensors 2022, 22(20), 7861; https://doi.org/10.3390/s22207861 - 16 Oct 2022

Cited by 6 | Viewed by 2004

Abstract

In direct line with the evolution of technology, but also with the density of vehicles that create congestion and often road accidents, traffic monitoring systems are parts that integrate intelligent transport systems (ITS). This is one of the most critical elements within transport [...] Read more.

In direct line with the evolution of technology, but also with the density of vehicles that create congestion and often road accidents, traffic monitoring systems are parts that integrate intelligent transport systems (ITS). This is one of the most critical elements within transport infrastructures, an aspect that involves extremely important financial investments in order to collect and analyze traffic data with the aim of designing systems capable of properly managing traffic. Technological progress in the field of wireless communications is advancing, highlighting new traffic monitoring solutions, and the need for major classification, but proposing a real-time analysis model to guide the new systems is a challenge addressed in this manuscript. The involvement of classifiers and computerized detection applied to traffic monitoring cameras can outline extremely vital systems for the future of logistic transport. Analyzing and debating vehicle classification systems, examining problems and challenges, as well as designing a software project capable of being the basis of new developments in the field of ITS systems are the aim of this study. The outline of a method based on intelligent algorithms and improved YOLOv3 can have a major impact on the effort to reduce the negative impact created by chaotic traffic and the outline of safety protocols in the field of transport. The reduction of waiting times and decongestion by up to 80% is a valid aspect, which we can deduce from the study carried out. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

24 pages, 2518 KiB

Open AccessArticle

RSU-Based Online Intrusion Detection and Mitigation for VANET

by Ammar Haydari and Yasin Yilmaz

Sensors 2022, 22(19), 7612; https://doi.org/10.3390/s22197612 - 08 Oct 2022

Cited by 14 | Viewed by 2567

Abstract

Secure vehicular communication is a critical factor for secure traffic management. Effective security in intelligent transportation systems (ITS) requires effective and timely intrusion detection systems (IDS). In this paper, we consider false data injection attacks and distributed denial-of-service (DDoS) attacks, especially the stealthy [...] Read more.

Secure vehicular communication is a critical factor for secure traffic management. Effective security in intelligent transportation systems (ITS) requires effective and timely intrusion detection systems (IDS). In this paper, we consider false data injection attacks and distributed denial-of-service (DDoS) attacks, especially the stealthy DDoS attacks, targeting integrity and availability, respectively, in vehicular ad-hoc networks (VANET). Novel machine learning techniques for intrusion detection and mitigation based on centralized communications through roadside units (RSU) are proposed for the considered attacks. The performance of the proposed methods is evaluated using a traffic simulator and a real traffic dataset. Comparisons with the state-of-the-art solutions clearly demonstrate the superior detection and localization performance of the proposed methods by

78 %

in the best case and

27 %

in the worst case, while achieving the same level of false alarm probability. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

19 pages, 8672 KiB

Open AccessArticle

Repetition-Based Approach for Task Adaptation in Imitation Learning

by Tho Nguyen Duc, Chanh Minh Tran, Nguyen Gia Bach, Phan Xuan Tan and Eiji Kamioka

Sensors 2022, 22(18), 6959; https://doi.org/10.3390/s22186959 - 14 Sep 2022

Viewed by 1216

Abstract

Transfer learning is an effective approach for adapting an autonomous agent to a new target task by transferring knowledge learned from the previously learned source task. The major problem with traditional transfer learning is that it only focuses on optimizing learning performance on [...] Read more.

Transfer learning is an effective approach for adapting an autonomous agent to a new target task by transferring knowledge learned from the previously learned source task. The major problem with traditional transfer learning is that it only focuses on optimizing learning performance on the target task. Thus, the performance on the target task may be improved in exchange for the deterioration of the source task’s performance, resulting in an agent that is not able to revisit the earlier task. Therefore, transfer learning methods are still far from being comparable with the learning capability of humans, as humans can perform well on both source and new target tasks. In order to address this limitation, a task adaptation method for imitation learning is proposed in this paper. Being inspired by the idea of repetition learning in neuroscience, the proposed adaptation method enables the agent to repeatedly review the learned knowledge of the source task, while learning the new knowledge of the target task. This ensures that the learning performance on the target task is high, while the deterioration of the learning performance on the source task is small. A comprehensive evaluation over several simulated tasks with varying difficulty levels shows that the proposed method can provide high and consistent performance on both source and target tasks, outperforming existing transfer learning methods. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

29 pages, 10934 KiB

Open AccessArticle

Predicting Bulk Average Velocity with Rigid Vegetation in Open Channels Using Tree-Based Machine Learning: A Novel Approach Using Explainable Artificial Intelligence

by D. P. P. Meddage, I. U. Ekanayake, Sumudu Herath, R. Gobirahavan, Nitin Muttil and Upaka Rathnayake

Sensors 2022, 22(12), 4398; https://doi.org/10.3390/s22124398 - 10 Jun 2022

Cited by 12 | Viewed by 2147

Abstract

Predicting the bulk-average velocity (U_B) in open channels with rigid vegetation is complicated due to the non-linear nature of the parameters. Despite their higher accuracy, existing regression models fail to highlight the feature importance or causality of the respective predictions. Therefore, [...] Read more.

Predicting the bulk-average velocity (U_B) in open channels with rigid vegetation is complicated due to the non-linear nature of the parameters. Despite their higher accuracy, existing regression models fail to highlight the feature importance or causality of the respective predictions. Therefore, we propose a method to predict U_B and the friction factor in the surface layer (f_S) using tree-based machine learning (ML) models (decision tree, extra tree, and XGBoost). Further, Shapley Additive exPlanation (SHAP) was used to interpret the ML predictions. The comparison emphasized that the XGBoost model is superior in predicting U_B (R = 0.984) and f_S (R = 0.92) relative to the existing regression models. SHAP revealed the underlying reasoning behind predictions, the dependence of predictions, and feature importance. Interestingly, SHAP adheres to what is generally observed in complex flow behavior, thus, improving trust in predictions. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

14 pages, 2383 KiB

Open AccessArticle

Smart Monitoring of Manufacturing Systems for Automated Decision-Making: A Multi-Method Framework

by Chen-Yang Cheng, Pourya Pourhejazy, Chia-Yu Hung and Chumpol Yuangyai

Sensors 2021, 21(20), 6860; https://doi.org/10.3390/s21206860 - 15 Oct 2021

Cited by 4 | Viewed by 2660

Abstract

Smart monitoring plays a principal role in the intelligent automation of manufacturing systems. Advanced data collection technologies, like sensors, have been widely used to facilitate real-time data collection. Computationally efficient analysis of the operating systems, however, remains relatively underdeveloped and requires more attention. [...] Read more.

Smart monitoring plays a principal role in the intelligent automation of manufacturing systems. Advanced data collection technologies, like sensors, have been widely used to facilitate real-time data collection. Computationally efficient analysis of the operating systems, however, remains relatively underdeveloped and requires more attention. Inspired by the capabilities of signal analysis and information visualization, this study proposes a multi-method framework for the smart monitoring of manufacturing systems and intelligent decision-making. The proposed framework uses the machine signals collected by noninvasive sensors for processing. For this purpose, the signals are filtered and classified to facilitate the realization of the operational status and performance measures to advise the appropriate course of managerial actions considering the detected anomalies. Numerical experiments based on real data are used to show the practicability of the developed monitoring framework. Results are supportive of the accuracy of the method. Applications of the developed approach are worthwhile research topics to research in other manufacturing environments. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

17 pages, 2934 KiB

Open AccessArticle

Enhanced Changeover Detection in Industry 4.0 Environments with Machine Learning

by Eddi Miller, Vladyslav Borysenko, Moritz Heusinger, Niklas Niedner, Bastian Engelmann and Jan Schmitt

Sensors 2021, 21(17), 5896; https://doi.org/10.3390/s21175896 - 01 Sep 2021

Cited by 6 | Viewed by 2567

Abstract

Changeover times are an important element when evaluating the Overall Equipment Effectiveness (OEE) of a production machine. The article presents a machine learning (ML) approach that is based on an external sensor setup to automatically detect changeovers in a shopfloor environment. The door [...] Read more.

Changeover times are an important element when evaluating the Overall Equipment Effectiveness (OEE) of a production machine. The article presents a machine learning (ML) approach that is based on an external sensor setup to automatically detect changeovers in a shopfloor environment. The door statuses, coolant flow, power consumption, and operator indoor GPS data of a milling machine were used in the ML approach. As ML methods, Decision Trees, Support Vector Machines, (Balanced) Random Forest algorithms, and Neural Networks were chosen, and their performance was compared. The best results were achieved with the Random Forest ML model (97% F1 score, 99.72% AUC score). It was also carried out that model performance is optimal when only a binary classification of a changeover phase and a production phase is considered and less subphases of the changeover process are applied. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

14 pages, 11142 KiB

Open AccessArticle

Domain Adaptation for Imitation Learning Using Generative Adversarial Network

by Tho Nguyen Duc, Chanh Minh Tran, Phan Xuan Tan and Eiji Kamioka

Sensors 2021, 21(14), 4718; https://doi.org/10.3390/s21144718 - 09 Jul 2021

Cited by 1 | Viewed by 2560

Abstract

Imitation learning is an effective approach for an autonomous agent to learn control policies when an explicit reward function is unavailable, using demonstrations provided from an expert. However, standard imitation learning methods assume that the agents and the demonstrations provided by the expert [...] Read more.

Imitation learning is an effective approach for an autonomous agent to learn control policies when an explicit reward function is unavailable, using demonstrations provided from an expert. However, standard imitation learning methods assume that the agents and the demonstrations provided by the expert are in the same domain configuration. Such an assumption has made the learned policies difficult to apply in another distinct domain. The problem is formalized as domain adaptive imitation learning, which is the process of learning how to perform a task optimally in a learner domain, given demonstrations of the task in a distinct expert domain. We address the problem by proposing a model based on Generative Adversarial Network. The model aims to learn both domain-shared and domain-specific features and utilizes it to find an optimal policy across domains. The experimental results show the effectiveness of our model in a number of tasks ranging from low to complex high-dimensional. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

22 pages, 24735 KiB

Open AccessArticle

A Few-Shot U-Net Deep Learning Model for COVID-19 Infected Area Segmentation in CT Images

by Athanasios Voulodimos, Eftychios Protopapadakis, Iason Katsamenis, Anastasios Doulamis and Nikolaos Doulamis

Sensors 2021, 21(6), 2215; https://doi.org/10.3390/s21062215 - 22 Mar 2021

Cited by 65 | Viewed by 5568

Abstract

Recent studies indicate that detecting radiographic patterns on CT chest scans can yield high sensitivity and specificity for COVID-19 identification. In this paper, we scrutinize the effectiveness of deep learning models for semantic segmentation of pneumonia-infected area segmentation in CT images for the [...] Read more.

Recent studies indicate that detecting radiographic patterns on CT chest scans can yield high sensitivity and specificity for COVID-19 identification. In this paper, we scrutinize the effectiveness of deep learning models for semantic segmentation of pneumonia-infected area segmentation in CT images for the detection of COVID-19. Traditional methods for CT scan segmentation exploit a supervised learning paradigm, so they (a) require large volumes of data for their training, and (b) assume fixed (static) network weights once the training procedure has been completed. Recently, to overcome these difficulties, few-shot learning (FSL) has been introduced as a general concept of network model training using a very small amount of samples. In this paper, we explore the efficacy of few-shot learning in U-Net architectures, allowing for a dynamic fine-tuning of the network weights as new few samples are being fed into the U-Net. Experimental results indicate improvement in the segmentation accuracy of identifying COVID-19 infected regions. In particular, using 4-fold cross-validation results of the different classifiers, we observed an improvement of 5.388 ± 3.046% for all test data regarding the IoU metric and a similar increment of 5.394 ± 3.015% for the F1 score. Moreover, the statistical significance of the improvement obtained using our proposed few-shot U-Net architecture compared with the traditional U-Net model was confirmed by applying the Kruskal-Wallis test (p-value = 0.026). Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Graphical abstract

20 pages, 9244 KiB

Open AccessArticle

Real-Time Rainfall Forecasts Based on Radar Reflectivity during Typhoons: Case Study in Southeastern Taiwan

by Chih-Chiang Wei and Chen-Chia Hsu

Sensors 2021, 21(4), 1421; https://doi.org/10.3390/s21041421 - 18 Feb 2021

Cited by 5 | Viewed by 2196

Abstract

This study developed a real-time rainfall forecasting system that can predict rainfall in a particular area a few hours before a typhoon’s arrival. The reflectivity of nine elevation angles obtained from the volume coverage pattern 21 Doppler radar scanning strategy and ground-weather data [...] Read more.

This study developed a real-time rainfall forecasting system that can predict rainfall in a particular area a few hours before a typhoon’s arrival. The reflectivity of nine elevation angles obtained from the volume coverage pattern 21 Doppler radar scanning strategy and ground-weather data of a specific area were used for accurate rainfall prediction. During rainfall prediction and analysis, rainfall retrievals were first performed to select the optimal radar scanning elevation angle for rainfall prediction at the current time. Subsequently, forecasting models were established using a single reflectivity and all elevation angles (10 prediction submodels in total) to jointly predict real-time rainfall and determine the optimal predicted values. This study was conducted in southeastern Taiwan and included three onshore weather stations (Chenggong, Taitung, and Dawu) and one offshore weather station (Lanyu). Radar reflectivities were collected from Hualien weather surveillance radar. The data for a total of 14 typhoons that affected the study area in 2008–2017 were collected. The gated recurrent unit (GRU) neural network was used to establish the forecasting model, and extreme gradient boosting and multiple linear regression were used as the benchmarks. Typhoons Nepartak, Meranti, and Megi were selected for simulation. The results revealed that the input data set merged with weather-station data, and radar reflectivity at the optimal elevation angle yielded optimal results for short-term rainfall forecasting. Moreover, the GRU neural network can obtain accurate predictions 1, 3, and 6 h before typhoon occurrence. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

13 pages, 2832 KiB

Open AccessArticle

Laser Cut Interruption Detection from Small Images by Using Convolutional Neural Network

by Benedikt Adelmann, Max Schleier and Ralf Hellmann

Sensors 2021, 21(2), 655; https://doi.org/10.3390/s21020655 - 19 Jan 2021

Cited by 6 | Viewed by 3179

Abstract

In this publication, we use a small convolutional neural network to detect cut interruptions during laser cutting from single images of a high-speed camera. A camera takes images without additional illumination at a resolution of 32 × 64 pixels from cutting steel sheets [...] Read more.

In this publication, we use a small convolutional neural network to detect cut interruptions during laser cutting from single images of a high-speed camera. A camera takes images without additional illumination at a resolution of 32 × 64 pixels from cutting steel sheets of varying thicknesses with different laser parameter combinations and classifies them into cuts and cut interruptions. After a short learning period of five epochs on a certain sheet thickness, the images are classified with a low error rate of 0.05%. The use of color images reveals slight advantages with lower error rates over greyscale images, since, during cut interruptions, the image color changes towards blue. A training set on all sheet thicknesses in one network results in tests error rates below 0.1%. This low error rate and the short calculation time of 120 µs on a standard CPU makes the system industrially applicable. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

21 pages, 1740 KiB

Open AccessArticle

Estimating Occupancy Levels in Enclosed Spaces Using Environmental Variables: A Fitness Gym and Living Room as Evaluation Scenarios

by Andree Vela, Joanna Alvarado-Uribe, Manuel Davila, Neil Hernandez-Gress and Hector G. Ceballos

Sensors 2020, 20(22), 6579; https://doi.org/10.3390/s20226579 - 18 Nov 2020

Cited by 15 | Viewed by 3310

Abstract

The understanding of occupancy patterns has been identified as a key contributor to achieve improvements in energy efficiency in buildings since occupancy information can benefit different systems, such as HVAC (Heating, Ventilation, and Air Conditioners), lighting, security, and emergency. This has meant that [...] Read more.

The understanding of occupancy patterns has been identified as a key contributor to achieve improvements in energy efficiency in buildings since occupancy information can benefit different systems, such as HVAC (Heating, Ventilation, and Air Conditioners), lighting, security, and emergency. This has meant that in the past decade, researchers have focused on improving the precision of occupancy estimation in enclosed spaces. Although several works have been done, one of the less addressed issues, regarding occupancy research, has been the availability of data for contrasting experimental results. Therefore, the main contributions of this work are: (1) the generation of two robust datasets gathered in enclosed spaces (a fitness gym and a living room) labeled with occupancy levels, and (2) the evaluation of three Machine Learning algorithms using different temporal resolutions. The results show that the prediction of 3–4 occupancy levels using the temperature, humidity, and pressure values provides an accuracy of at least 97%. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

19 pages, 2347 KiB

Open AccessArticle

Interterminal Truck Routing Optimization Using Deep Reinforcement Learning

by Taufik Nur Adi, Yelita Anggiane Iskandar and Hyerim Bae

Sensors 2020, 20(20), 5794; https://doi.org/10.3390/s20205794 - 13 Oct 2020

Cited by 19 | Viewed by 4199

Abstract

The continued growth of the volume of global containerized transport necessitates that most of the major ports in the world improve port productivity by investing in more interconnected terminals. The development of the multiterminal system escalates the complexity of the container transport process [...] Read more.

The continued growth of the volume of global containerized transport necessitates that most of the major ports in the world improve port productivity by investing in more interconnected terminals. The development of the multiterminal system escalates the complexity of the container transport process and increases the demand for container exchange between different terminals within a port, known as interterminal transport (ITT). Trucks are still the primary modes of freight transportation to transport containers among most terminals. A trucking company needs to consider proper truck routing planning because, based on several studies, it played an essential role in coordinating ITT flows. Furthermore, optimal truck routing in the context of ITT significantly affects port productivity and efficiency. The study of deep reinforcement learning in truck routing optimization is still limited. In this study, we propose deep reinforcement learning to provide truck routes of a given container transport order by considering several significant factors such as order origin, destination, time window, and due date. To assess its performance, we compared between the proposed method and two approaches that are used to solve truck routing problems. The experiment results showed that the proposed method obtains considerably better results compared to the other algorithms. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

23 pages, 11705 KiB

Open AccessArticle

Degradation Tendency Prediction for Pumped Storage Unit Based on Integrated Degradation Index Construction and Hybrid CNN-LSTM Model

by Jianzhong Zhou, Yahui Shan, Jie Liu, Yanhe Xu and Yang Zheng

Sensors 2020, 20(15), 4277; https://doi.org/10.3390/s20154277 - 31 Jul 2020

Cited by 18 | Viewed by 3060

Abstract

Accurate degradation tendency prediction (DTP) is vital for the secure operation of a pumped storage unit (PSU). However, the existing techniques and methodologies for DTP still face challenges, such as a lack of appropriate degradation indicators, insufficient accuracy, and poor capability to track [...] Read more.

Accurate degradation tendency prediction (DTP) is vital for the secure operation of a pumped storage unit (PSU). However, the existing techniques and methodologies for DTP still face challenges, such as a lack of appropriate degradation indicators, insufficient accuracy, and poor capability to track the data fluctuation. In this paper, a hybrid model is proposed for the degradation tendency prediction of a PSU, which combines the integrated degradation index (IDI) construction and convolutional neural network-long short-term memory (CNN-LSTM). Firstly, the health model of a PSU is constructed with Gaussian process regression (GPR) and the condition parameters of active power, working head, and guide vane opening. Subsequently, for comprehensively quantifying the degradation level of PSU, an IDI is developed using entropy weight (EW) theory. Finally, combining the local feature extraction of the CNN with the time series representation of LSTM, the CNN-LSTM model is constructed to realize DTP. To validate the effectiveness of the proposed model, the monitoring data collected from a PSU in China is taken as case studies. The root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) obtained by the proposed model are 1.1588, 0.8994, 0.0918, and 0.9713, which can meet the engineering application requirements. The experimental results show that the proposed model outperforms other comparison models. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

Review

Jump to: Research

33 pages, 5318 KiB

Open AccessReview

A Review on Machine Learning Applications for Solar Plants

by Ekaterina Engel and Nikita Engel

Sensors 2022, 22(23), 9060; https://doi.org/10.3390/s22239060 - 22 Nov 2022

Cited by 3 | Viewed by 2221

Abstract

A solar plant system has complex nonlinear dynamics with uncertainties due to variations in system parameters and insolation. Thereby, it is difficult to approximate these complex dynamics with conventional algorithms whereas Machine Learning (ML) methods yield the essential performance required. ML models are [...] Read more.

A solar plant system has complex nonlinear dynamics with uncertainties due to variations in system parameters and insolation. Thereby, it is difficult to approximate these complex dynamics with conventional algorithms whereas Machine Learning (ML) methods yield the essential performance required. ML models are key units in recent sensor systems for solar plant design, forecasting, maintenance, and control to provide the best safety, reliability, robustness, and performance as compared to classical methods which are usually employed in the hardware and software of solar plants. Considering this, the goal of our paper is to explore and analyze ML technologies and their advantages and shortcomings as compared to classical methods for the design, forecasting, maintenance, and control of solar plants. In contrast with other review articles, our research briefly summarizes our intelligent, self-adaptive models for sizing, forecasting, maintenance, and control of a solar plant; sets benchmarks for performance comparison of the reviewed ML models for a solar plant’s system; proposes a simple but effective integration scheme of an ML sensor solar plant system’s implementation and outlines its future digital transformation into a smart solar plant based on the integrated cutting-edge technologies; and estimates the impact of ML technologies based on the proposed scheme on a solar plant value chain. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

33 pages, 1844 KiB

Open AccessReview

Seven-Layer Model in Complex Networks Link Prediction: A Survey

by Hui Wang and Zichun Le

Sensors 2020, 20(22), 6560; https://doi.org/10.3390/s20226560 - 17 Nov 2020

Cited by 16 | Viewed by 3096

Abstract

Link prediction is the most basic and essential problem in complex networks. This study analyzes the observed topological, time, attributive, label, weight, directional, and symbolic features and auxiliary information to find the lack of connection and predict the future possible connection. For discussion [...] Read more.

Link prediction is the most basic and essential problem in complex networks. This study analyzes the observed topological, time, attributive, label, weight, directional, and symbolic features and auxiliary information to find the lack of connection and predict the future possible connection. For discussion and analysis of the evolution of the network, the network model is of great significance. In the past two decades, link prediction has attracted extensive attention from experts in various fields, who have published numerous high-level papers, but few combine interdisciplinary characteristics. This survey analyzes and discusses the existing link prediction methods. The idea of stratification is introduced into the classification system of link prediction for the first time and proposes the design idea of a seven-layer model, namely the network, metadata, feature classification, selection input, processing, selection, and output layers. Among them, the processing layer divides link prediction methods into similarity-based, probabilistic, likelihood, supervised learning, semi-supervised learning, unsupervised learning, and reinforcement learning methods. The input features, evaluation metrics, complex analysis, experimental comparisons, relative merits, common dataset and open-source implementations for each link prediction method are then discussed in detail. Through analysis and comparison, we found that the link prediction method based on graph structure features has better prediction performance. Finally, the future development direction of link prediction in complex networks is discussed. Full article

(This article belongs to the Special Issue Advances in Machine Learning for Intelligent Engineering Systems and Applications II)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Advances in Machine Learning for Intelligent Engineering Systems and Applications II

Share This Special Issue

Special Issue Editors

Special Issue Information

Keywords

Related Special Issue

Published Papers (23 papers)

Research

Review

Further Information

Guidelines

MDPI Initiatives

Follow MDPI