mdh.sePublications
Change search
Link to record
Permanent link

Direct link
BETA
Publications (10 of 28) Show all publications
Rahman, H., Ahmed, M. U., Barua, S. & Begum, S. (2020). Non-contact-based driver's cognitive load classification using physiological and vehicular parameters. Biomedical Signal Processing and Control, 55, Article ID 101634.
Open this publication in new window or tab >>Non-contact-based driver's cognitive load classification using physiological and vehicular parameters
2020 (English)In: Biomedical Signal Processing and Control, ISSN 1746-8094, E-ISSN 1746-8108, Vol. 55, article id 101634Article in journal (Refereed) Published
Abstract [en]

Classification of cognitive load for vehicular drivers is a complex task due to underlying challenges of the dynamic driving environment. Many previous works have shown that physiological sensor signals or vehicular data could be a reliable source to quantify cognitive load. However, in driving situations, one of the biggest challenges is to use a sensor source that can provide accurate information without interrupting diverging tasks. In this paper, instead of traditional wire-based sensors, non-contact camera and vehicle data are used that have no physical contact with the driver and do not interrupt driving. Here, four machine learning algorithms, logistic regression (LR), support vector machine (SVM), linear discriminant analysis (LDA) and neural networks (NN), are investigated to classify the cognitive load using the collected data from a driving simulator study. In this paper, physiological parameters are extracted from facial video images, and vehicular parameters are collected from controller area networks (CAN). The data collection was performed in close collaboration with industrial partners in two separate studies, in which study-1 was designed with a 1-back task and study-2 was designed with both 1-back and 2-back task. The goal of the experiment is to investigate how accurately the machine learning algorithms can classify drivers' cognitive load based on the extracted features in complex dynamic driving environments. According to the results, for the physiological parameters extracted from the facial videos, the LR model with logistic function outperforms the other three classification methods. Here, in study-1, the achieved average accuracy for the LR classifier is 94% and in study-2 the average accuracy is 82%. In addition, the classification accuracy for the collected physiological parameters was compared with reference wire-sensor signals. It is observed that the classification accuracies between the sensor and the camera are very similar; however, better accuracy is achieved with the camera data due to having lower artefacts than the sensor data. 

Place, publisher, year, edition, pages
ELSEVIER SCI LTD, 2020
Keywords
Non-contact, Physiological parameters, Vehicular parameters, Cognitive load, Classification, Logistic regression, Support vector machine, Decision tree
National Category
Computer Systems
Identifiers
urn:nbn:se:mdh:diva-46634 (URN)10.1016/j.bspc.2019.101634 (DOI)000502893200022 ()2-s2.0-85071533851 (Scopus ID)
Available from: 2020-01-02 Created: 2020-01-02 Last updated: 2020-02-20Bibliographically approved
Ahmed, M. U., Andersson, P., Andersson, T., Tomas Aparicio, E., Baaz, H., Barua, S., . . . Zambrano, J. (2019). A Machine Learning Approach for Biomass Characterization. In: Yan, J Yang, HX Li, H Chen, X (Ed.), INNOVATIVE SOLUTIONS FOR ENERGY TRANSITIONS: . Paper presented at 10th International Conference on Applied Energy (ICAE), AUG 22-25, 2018, Hong Kong, HONG KONG (pp. 1279-1287). ELSEVIER SCIENCE BV
Open this publication in new window or tab >>A Machine Learning Approach for Biomass Characterization
Show others...
2019 (English)In: INNOVATIVE SOLUTIONS FOR ENERGY TRANSITIONS / [ed] Yan, J Yang, HX Li, H Chen, X, ELSEVIER SCIENCE BV , 2019, p. 1279-1287Conference paper, Published paper (Refereed)
Abstract [en]

The aim of this work is to apply and evaluate different chemometric approaches employing several machine learning techniques in order to characterize the moisture content in biomass from data obtained by Near Infrared (NIR) spectroscopy. The approaches include three main parts: a) data pre-processing, b) wavelength selection and c) development of a regression model enabling moisture content measurement. Standard Normal Variate (SNV), Multiplicative Scatter Correction and Savitzky-Golay first (SGi) and second (SG2) derivatives and its combinations were applied for data pre-processing. Genetic algorithm (GA) and iterative PLS (iPLS) were used for wavelength selection. Artificial Neural Network (ANN), Gaussian Process Regression (GPR), Support Vector Regression (SVR) and traditional Partial Least Squares (PLS) regression, were employed as machine learning regression methods. Results shows that SNV combined with SG1 first derivative performs the best in data pre-processing. The GA is the most effective methods for variable selection and GPR achieved a high accuracy in regression modeling while having low demands on computation time. Overall, the machine learning techniques demonstrate a great potential to be used in future NIR spectroscopy applications.

Place, publisher, year, edition, pages
ELSEVIER SCIENCE BV, 2019
Series
Energy Procedia, ISSN 1876-6102 ; 158
Keywords
Artificial Neural Network, Chemometrics, Gaussian Process Regression, Near Infrared Spectroscopy, Multiplicative Scatter Correction, Standard Normal Variate, Support Vector Regression, Partial Least Squares, Savitzky-Golay derivatives
National Category
Energy Engineering
Identifiers
urn:nbn:se:mdh:diva-44835 (URN)10.1016/j.egypro.2019.01.316 (DOI)000471031701100 ()2-s2.0-85063865772 (Scopus ID)
Conference
10th International Conference on Applied Energy (ICAE), AUG 22-25, 2018, Hong Kong, HONG KONG
Available from: 2019-07-11 Created: 2019-07-11 Last updated: 2019-10-14Bibliographically approved
Barua, S., Ahmed, M. U., Ahlström, C. & Begum, S. (2019). Automatic driver sleepiness detection using EEG, EOG and contextual information. Expert systems with applications, 115, 121-135
Open this publication in new window or tab >>Automatic driver sleepiness detection using EEG, EOG and contextual information
2019 (English)In: Expert systems with applications, ISSN 0957-4174, E-ISSN 1873-6793, Vol. 115, p. 121-135Article in journal (Refereed) Published
Abstract [en]

The many vehicle crashes that are caused by driver sleepiness each year advocates the development of automated driver sleepiness detection (ADSD) systems. This study proposes an automatic sleepiness classification scheme designed using data from 30 drivers who repeatedly drove in a high-fidelity driving simulator, both in alert and in sleep deprived conditions. Driver sleepiness classification was performed using four separate classifiers: k-nearest neighbours, support vector machines, case-based reasoning, and random forest, where physiological signals and contextual information were used as sleepiness indicators. The subjective Karolinska sleepiness scale (KSS) was used as target value. An extensive evaluation on multiclass and binary classifications was carried out using 10-fold cross-validation and leave-one-out validation. With 10-fold cross-validation, the support vector machine showed better performance than the other classifiers (79% accuracy for multiclass and 93% accuracy for binary classification). The effect of individual differences was also investigated, showing a 10% increase in accuracy when data from the individual being evaluated was included in the training dataset. Overall, the support vector machine was found to be the most stable classifier. The effect of adding contextual information to the physiological features improved the classification accuracy by 4% in multiclass classification and by and 5% in binary classification.

Place, publisher, year, edition, pages
Elsevier Ltd, 2019
Keywords
Contextual information, Driver sleepiness, Electroencephalography, Electrooculography, Machine learning, Accidents, Case based reasoning, Decision trees, Electrophysiology, Fisher information matrix, Learning systems, Nearest neighbor search, Support vector machines, 10-fold cross-validation, Binary classification, Classification accuracy, Individual Differences, Multi-class classification, Physiological features, Classification (of information)
National Category
Electrical Engineering, Electronic Engineering, Information Engineering
Identifiers
urn:nbn:se:mdh:diva-40526 (URN)10.1016/j.eswa.2018.07.054 (DOI)000448097700009 ()2-s2.0-85051410923 (Scopus ID)
Available from: 2018-08-23 Created: 2018-08-23 Last updated: 2019-01-10Bibliographically approved
Islam, M. R., Barua, S., Ahmed, M. U., Begum, S. & Flumeri, G. D. (2019). Deep Learning for Automatic EEG Feature Extraction: An Application in Drivers' Mental Workload Classification. In: Communications in Computer and Information Science, Volume 1107: . Paper presented at The 3rd International Symposium on Human Mental Workload: Models and Applications H-WORKLOAD 2019, 14 Nov 2019, Rome, Italy (pp. 121-135).
Open this publication in new window or tab >>Deep Learning for Automatic EEG Feature Extraction: An Application in Drivers' Mental Workload Classification
Show others...
2019 (English)In: Communications in Computer and Information Science, Volume 1107, 2019, p. 121-135Conference paper, Published paper (Refereed)
Abstract [en]

In the pursuit of reducing traffic accidents, drivers' mental workload (MWL) has been considered as one of the vital aspects. To measure MWL in different driving situations Electroencephalography (EEG) of the drivers has been studied intensely. However, in the literature, mostly, manual analytic methods are applied to extract and select features from the EEG signals to quantify drivers' MWL. Nevertheless, the amount of time and effort required to perform prevailing feature extraction techniques leverage the need for automated feature extraction techniques. This work investigates deep learning (DL) algorithm to extract and select features from the EEG signals during naturalistic driving situations. Here, to compare the DL based and traditional feature extraction techniques, a number of classifiers have been deployed. Results have shown that the highest value of area under the curve of the receiver operating characteristic (AUC-ROC) is 0.94, achieved using the features extracted by CNN-AE and support vector machine. Whereas, using the features extracted by the traditional method, the highest value of AUC-ROC is 0.78 with the multi-layer perceptron. Thus, the outcome of this study shows that the automatic feature extraction techniques based on CNN-AE can outperform the manual techniques in terms of classification accuracy.

Keywords
Autoencoder, Convolutional Neural Networks, Electroencephalography, Feature Extraction, Mental Workload
National Category
Engineering and Technology Computer Systems
Identifiers
urn:nbn:se:mdh:diva-45059 (URN)10.1007/978-3-030-32423-0_8 (DOI)2-s2.0-85075680380 (Scopus ID)9783030324223 (ISBN)
Conference
The 3rd International Symposium on Human Mental Workload: Models and Applications H-WORKLOAD 2019, 14 Nov 2019, Rome, Italy
Projects
BRAINSAFEDRIVE: A Technology to detect Mental States During Drive for improving the Safety of the road
Available from: 2019-08-22 Created: 2019-08-22 Last updated: 2019-12-16Bibliographically approved
Islam, M. R., Barua, S., Begum, S. & Ahmed, M. U. (2019). Hypothyroid Disease Diagnosis with Causal Explanation using Case-based Reasoning and Domain-specific Ontology. In: Workshop on CBR in the Health Science WS-HealthCBR: . Paper presented at Workshop on CBR in the Health Science WS-HealthCBR, 09 Sep 2019, Otzenhausen, Germany.
Open this publication in new window or tab >>Hypothyroid Disease Diagnosis with Causal Explanation using Case-based Reasoning and Domain-specific Ontology
2019 (English)In: Workshop on CBR in the Health Science WS-HealthCBR, 2019Conference paper, Published paper (Refereed)
Abstract [en]

Explainability of intelligent systems in health-care domain is still in its initial state. Recently, more efforts are made to leverage machine learning in solving causal inference problems of disease diagnosis, prediction and treatments. This research work presents an ontology based causal inference model for hypothyroid disease diagnosis using case-based reasoning. The effectiveness of the proposed method is demonstrated with an example from hypothyroid disease domain. Here, the domain knowledge is mapped into an ontology and causal inference is performed based on this domain-specific ontology. The goal is to incorporate this causal inference model in traditional case-based reasoning cycle enabling explanation for each solved problem. Finally, a mechanism is defined to deduce explanation for a solution to a problem case from the combined causal statements of similar cases. The initial result shows that case-based reasoning can retrieve relevant cases with 95% accuracy.

Keywords
Case-based Reasoning, Causal Model, Explainability, Explainable Artificial Intelligence, Hypothyroid Diagnosis, Ontology
National Category
Engineering and Technology Computer Systems
Identifiers
urn:nbn:se:mdh:diva-45058 (URN)
Conference
Workshop on CBR in the Health Science WS-HealthCBR, 09 Sep 2019, Otzenhausen, Germany
Available from: 2019-08-22 Created: 2019-08-22 Last updated: 2019-08-22Bibliographically approved
Ahmed, M. U., Andersson, P., Andersson, T., Tomas Aparicio, E., Baaz, H., Barua, S., . . . Zambrano, J. (2019). Real-time Biomass Characterization in Energy Conversion Processes using Near Infrared Spectroscopy: A Machine Learning Approach. In: Elsevier (Ed.), “Innovative Solutions for Energy Transitions”: . Paper presented at International Conference on Applied Energy, 2018 (pp. 1279-1287). , 158
Open this publication in new window or tab >>Real-time Biomass Characterization in Energy Conversion Processes using Near Infrared Spectroscopy: A Machine Learning Approach
Show others...
2019 (English)In: “Innovative Solutions for Energy Transitions” / [ed] Elsevier, 2019, Vol. 158, p. 1279-1287Conference paper, Published paper (Refereed)
Abstract [en]

The aim of this work is to apply and evaluate different chemometric approaches employing several machine learning techniques in order to characterize the moisture content in biomass from data obtained by Near Infrared (NIR) spectroscopy. The approaches include three main parts: a) data pre-processing, b) wavelength selection and c) development of a regression model enabling moisture content measurement. Standard Normal Variate (SNV), Multiplicative Scatter Correction and Savitzky-Golay first (SG1) and second (SG2) derivatives and its combinations were applied for data pre-processing. Genetic algorithm (GA) and iterative PLS (iPLS) were used for wavelength selection. Artificial Neural Network (ANN), Gaussian Process Regression (GPR), Support Vector Regression (SVR) and traditional Partial Least Squares (PLS) regression, were employed as machine learning regression methods. Results shows that SNV combined with SG1 first derivative performs the best in data pre-processing. The GA is the most effective methods for variable selection and GPR achieved a high accuracy in regression modeling while having low demands on computation time. Overall, the machine learning techniques demonstrate a great potential to be used in future NIR spectroscopy applications.

Series
Energy Procedia, ISSN 1876-6102
Keywords
Artificial Neural Netwrok; Chemometrics; Gaussian Process Regression; Multiplicative Scatter Correction; Standard Normal Variate; Support Vector Regression; Partial Least Squares; Savitzky-Golay derivatives
National Category
Environmental Engineering Environmental Biotechnology Industrial Biotechnology
Research subject
Energy- and Environmental Engineering
Identifiers
urn:nbn:se:mdh:diva-40396 (URN)10.1016/j.egypro.2019.01.316 (DOI)471031701100 ()2-s2.0-85063865772 (Scopus ID)
Conference
International Conference on Applied Energy, 2018
Projects
FUDIPO
Funder
EU, Horizon 2020, 723523
Available from: 2018-08-21 Created: 2018-08-21 Last updated: 2020-03-16Bibliographically approved
Barua, S., Ahmed, M. U. & Begum, S. (2018). Distributed Multivariate Physiological Signal Analytics for Driver´s Mental State Monitoring. In: Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST, Volume 225: . Paper presented at 4th EAI International Conference on IoT Technologies for HealthCare HealthyIOT'17, 24 Oct 2017, Angers, France (pp. 26-33).
Open this publication in new window or tab >>Distributed Multivariate Physiological Signal Analytics for Driver´s Mental State Monitoring
2018 (English)In: Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST, Volume 225, 2018, p. 26-33Conference paper, Published paper (Refereed)
Abstract [en]

This paper presents a distributed data analytics approach for drivers’ mental state monitoring using multivariate physiological signals. Driver’s mental states such as cognitive distraction, sleepiness, stress, etc. can be fatal contributing factors and to prevent car crashes these factors need to be understood. Here, a cloud-based approach with heterogeneous sensor sources that generates extremely large data sets of physiological signals need to be handled and analyzed in a big data scenario. In the proposed physiological big data analytics approach, for driver state monitoring, heterogeneous data coming from multiple sources i.e., multivariate physiological signals are used, processed and analyzed to aware impaired vehicle drivers. Here, in a distributed big data environment, multi-agent case-based reasoning facilitates parallel case similarity matching and handles data that are coming from single and multiple physiological signal sources.

Keywords
Physiological signals, distributed analytics, case-based reasoning
National Category
Medical Engineering
Identifiers
urn:nbn:se:mdh:diva-37076 (URN)10.1007/978-3-319-76213-5_4 (DOI)000476922000004 ()2-s2.0-85042522774 (Scopus ID)9783319762128 (ISBN)
Conference
4th EAI International Conference on IoT Technologies for HealthCare HealthyIOT'17, 24 Oct 2017, Angers, France
Projects
VDM - Vehicle Driver Monitoring
Available from: 2017-10-31 Created: 2017-10-31 Last updated: 2019-08-08Bibliographically approved
Barua, S., Begum, S. & Ahmed, M. U. (2018). Scalable Framework for Distributed Case-based Reasoning for Big data analytics. In: Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST, Volume 225: . Paper presented at 4th EAI International Conference on IoT Technologies for HealthCare HealthyIOT'17, 24 Oct 2017, Angers, France (pp. 111-114).
Open this publication in new window or tab >>Scalable Framework for Distributed Case-based Reasoning for Big data analytics
2018 (English)In: Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering, LNICST, Volume 225, 2018, p. 111-114Conference paper, Published paper (Refereed)
Abstract [en]

This paper proposes a scalable framework for distributed case-based reasoning methodology to provide actionable knowledge based on historical big amount of data. The framework addresses several challenges, i.e., promptly analyse big data, cross-domain, use-case specific data processing, multi-source case representation, dynamic case-management, uncertainty, check the plausibility of solution after adaptation etc. through its’ five modules architectures. The architecture allows the functionalities with distributed data analytics and intended to provide solutions under different conditions, i.e. data size, velocity, variety etc.

Keywords
Distributed analytics, case-based reasoning
National Category
Computer Systems
Identifiers
urn:nbn:se:mdh:diva-37087 (URN)10.1007/978-3-319-76213-5_16 (DOI)000476922000016 ()2-s2.0-85042527811 (Scopus ID)9783319762128 (ISBN)
Conference
4th EAI International Conference on IoT Technologies for HealthCare HealthyIOT'17, 24 Oct 2017, Angers, France
Projects
VDM - Vehicle Driver Monitoring
Available from: 2017-10-27 Created: 2017-10-27 Last updated: 2019-08-08Bibliographically approved
Barua, S., Begum, S. & Ahmed, M. U. (2018). Towards Distributed k-NN similarity for Scalable Case Retrieval. In: ICCBR 2018: The 26th International Conference on Case-Based Reasoning July, 09th-12th 2018 in Stockholm, Sweden, Workshop Proceedings. Paper presented at XCBR: First Workshop on Case-Based Reasoning for the Explanation of Intelligent Systems. Workshop at the 26th International Conference on Case-Based Reasoning (ICCBR 2018) (pp. 151-160).
Open this publication in new window or tab >>Towards Distributed k-NN similarity for Scalable Case Retrieval
2018 (English)In: ICCBR 2018: The 26th International Conference on Case-Based Reasoning July, 09th-12th 2018 in Stockholm, Sweden, Workshop Proceedings, 2018, p. 151-160Conference paper, Published paper (Refereed)
Abstract [en]

In Big data era, the demand of processing large amount of data posing several challenges. One biggest challenge is that it is no longer possible to process the data in a single machine. Similar challenges can be assumed for case-based reasoning (CBR) approach, where the size of a case library is increasing and constructed using heterogenous data sources. To deal with the challenges of big data in CBR, a distributed CBR system can be developed, where case libraries or cases are distributed over clusters. MapReduce programming framework has the facilities of parallel processing massive amount of data through a distributed system. This paper proposes a scalable case-representation and retrieval approach using distributed k-NN similarity. The proposed approach is considered to be developed using MapReduce programming framework, where cases are distributed in many clusters.

Keywords
Case representation, k-NN similarity, Scalable Case Retrieval, distributed CBR, big data
National Category
Computer Systems
Identifiers
urn:nbn:se:mdh:diva-40882 (URN)
Conference
XCBR: First Workshop on Case-Based Reasoning for the Explanation of Intelligent Systems. Workshop at the 26th International Conference on Case-Based Reasoning (ICCBR 2018)
Available from: 2018-09-18 Created: 2018-09-18 Last updated: 2018-09-18Bibliographically approved
Barua, S., Ahmed, M. U., Ahlström, C., Begum, S. & Funk, P. (2017). Automated EEG Artifact Handling with Application in Driver Monitoring. IEEE journal of biomedical and health informatics, 22(5), 1350-1361
Open this publication in new window or tab >>Automated EEG Artifact Handling with Application in Driver Monitoring
Show others...
2017 (English)In: IEEE journal of biomedical and health informatics, ISSN 2168-2194, E-ISSN 2168-2208, Vol. 22, no 5, p. 1350-1361Article in journal (Refereed) Published
Abstract [en]

Automated analyses of electroencephalographic (EEG) signals acquired in naturalistic environments is becoming increasingly important in areas such as brain computer interfaces and behaviour science. However, the recorded EEG in such environments is often heavily contaminated by motion artifacts and eye movements. This poses new requirements on artifact handling. The objective of this paper is to present an automated EEG artifacts handling algorithm which will be used as a pre-processing step in a driver monitoring application. The algorithm, named ARTE (Automated aRTifacts handling in EEG), is based on wavelets, independent component analysis and hierarchical clustering. The algorithm is tested on a dataset obtained from a driver sleepiness study including 30 drivers and 540 30-minute 30-channel EEG recordings. The algorithm is evaluated by a clinical neurophysiologist, by quantitative criteria (signal quality index, mean square error, relative error and mean absolute error), and by demonstrating its usefulness as a preprocessing step in driver monitoring, here exemplified with driver sleepiness classification. All results are compared with a state of the art algorithm called FORCe. The quantitative and expert evaluation results show that the two algorithms are comparable and that both algorithms significantly reduce the impact of artifacts in recorded EEG signals. When artifact handling is used as a pre-processing step in driver sleepiness classification, the classification accuracy increased by 5% when using ARTE and by 2% when using FORCe. The advantage with ARTE is that it is data driven and does not rely on additional reference signals or manually defined thresholds, making it well suited for use in dynamic settings where unforeseen and rare artifacts are commonly encountered.

Place, publisher, year, edition, pages
IEEE, 2017
Keywords
Artifacts, Clustering, Electroencephalogram, Independent Component Analysis, Wavelet decomposition
National Category
Signal Processing
Identifiers
urn:nbn:se:mdh:diva-37347 (URN)10.1109/JBHI.2017.2773999 (DOI)000441795800003 ()2-s2.0-85035807991 (Scopus ID)
Projects
VDM - Vehicle Driver MonitoringSafeDriver: A Real Time Driver's State Monitoring and Prediction System
Available from: 2017-11-27 Created: 2017-11-27 Last updated: 2019-01-10Bibliographically approved
Organisations
Identifiers
ORCID iD: ORCID iD iconorcid.org/0000-0002-7305-7169

Search in DiVA

Show all publications