Most recent issue published online in the International Journal of Intelligent Systems Technologies and Applications.
International Journal of Intelligent Systems Technologies and Applications
http://www.inderscience.com/browse/index.php?journalID=35&year=2024&vol=22&issue=1
Inderscience Publishers Ltd
en-uk
support@inderscience.com
International Journal of Intelligent Systems Technologies and Applications
1740-8865
1740-8873
© 2024 Inderscience Enterprises Ltd.
© 2024 Inderscience Publishers Ltd
editor@inderscience.com
International Journal of Intelligent Systems Technologies and Applications
https://www.inderscience.com/images/files/coverImgs/ijista_scoverijista.jpg
http://www.inderscience.com/browse/index.php?journalID=35&year=2024&vol=22&issue=1
-
Service capability aware big data workflow scheduling approach in cloud datacentre
http://www.inderscience.com/link.php?id=136520
With the increasing application of cloud computing, big data workflow scheduling in cloud datacentre also become an important focus of research. How to guarantee minimal scheduling length is the main challenge in scheduling workflow in cloud-based environments. The main limitation of proposed approaches stems is that they overlook the service capability support levels of the virtual machines and service capability requirement levels of the different tasks in a workflow, thus risking resulting in extremely poor processing efficiency. We propose a service dynamic level scheduling algorithm in cloud datacentre (Cloud-SDLS) that consists of three stages: virtual machines' service capability support computation, tasks' service capability requirement computation, and service dynamic level scheduling. Experimental results show that the proposed algorithms effectively satisfy the QoS in service capability requirement. It is significant to shorten workflow completion time in practice.
Service capability aware big data workflow scheduling approach in cloud datacentre
Jie Cao; Jinchao Xu; Bo Wang
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 1 - 15
With the increasing application of cloud computing, big data workflow scheduling in cloud datacentre also become an important focus of research. How to guarantee minimal scheduling length is the main challenge in scheduling workflow in cloud-based environments. The main limitation of proposed approaches stems is that they overlook the service capability support levels of the virtual machines and service capability requirement levels of the different tasks in a workflow, thus risking resulting in extremely poor processing efficiency. We propose a service dynamic level scheduling algorithm in cloud datacentre (Cloud-SDLS) that consists of three stages: virtual machines' service capability support computation, tasks' service capability requirement computation, and service dynamic level scheduling. Experimental results show that the proposed algorithms effectively satisfy the QoS in service capability requirement. It is significant to shorten workflow completion time in practice.]]>
10.1504/IJISTA.2024.136520
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 1 - 15
Jie Cao
Jinchao Xu
Bo Wang
Software Engineering College, Zhengzhou University of Light Industry, Zhengzhou, 450002, China ' Information Centre, Shanghai Jiaotong University, Shanghai, 200240, China ' Software Engineering College, Zhengzhou University of Light Industry, Zhengzhou, 450002, China
cloud computing
service capability requirement
service capability support
workflow scheduling
2024-02-05T23:20:50-05:00
Copyright © 2024 Inderscience Enterprises Ltd.
22
1
1
15
2024-02-05T23:20:50-05:00
-
Improving assaulted medical image quality using improved adaptive filtering network
http://www.inderscience.com/link.php?id=136531
Regular static convolutions work well for low-frequency information processing but fall short for high-frequency information processing. Dynamic convolution is the recent method that has spatial anisotropy and content-adaptiveness, enabling it to restore complicated and sensitive high-frequency information. The proposed method makes use of dynamic convolution to enhance the learning of multi-scale and high-frequency features. To accomplish this, two blocks - the dynamic convolution block (DCB) and the multi-scale dynamic convolution block (MDCB) are introduced. Dynamic convolution is used by the DCB to improve high-frequency information, whereas skip connections are used to protect low-frequency information. To efficiently extract multi-scale features, the MDCB uses shared adaptive dynamic kernels of increasing size along with dynamic convolution. The proposed multi-dimension feature integration mechanism is used to produce accurate and contextually enriched feature representations. For successful denoising, an improved adaptive dynamic filtering network is useful.
Improving assaulted medical image quality using improved adaptive filtering network
Namita D. Pulgam; Subhash K. Shinde
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 16 - 28
Regular static convolutions work well for low-frequency information processing but fall short for high-frequency information processing. Dynamic convolution is the recent method that has spatial anisotropy and content-adaptiveness, enabling it to restore complicated and sensitive high-frequency information. The proposed method makes use of dynamic convolution to enhance the learning of multi-scale and high-frequency features. To accomplish this, two blocks - the dynamic convolution block (DCB) and the multi-scale dynamic convolution block (MDCB) are introduced. Dynamic convolution is used by the DCB to improve high-frequency information, whereas skip connections are used to protect low-frequency information. To efficiently extract multi-scale features, the MDCB uses shared adaptive dynamic kernels of increasing size along with dynamic convolution. The proposed multi-dimension feature integration mechanism is used to produce accurate and contextually enriched feature representations. For successful denoising, an improved adaptive dynamic filtering network is useful.]]>
10.1504/IJISTA.2024.136531
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 16 - 28
Namita D. Pulgam
Subhash K. Shinde
Department of Computer Engineering, Ramrao Adik Institute of Technology, DY Patil (Deemed to be University), Navi Mumbai, India ' Department of Computer Engineering, Lokmanya Tilak College of Engineering, Navi Mumbai, India
image processing
medical image
digital watermarking
encryption
data security
denoising
deep convolutional neural network
2024-02-05T23:20:50-05:00
Copyright © 2024 Inderscience Enterprises Ltd.
22
1
16
28
2024-02-05T23:20:50-05:00
-
Developing a method to detect driver drowsiness based on a single EEG channel and discriminated features
http://www.inderscience.com/link.php?id=136521
Driver drowsiness is one of the leading causes of road deaths and transportation industry dangers. Due to its direct evaluation of neurophysiological brain activity, electroencephalography (EEG) has been regarded as one of the most reliable physiological indicators for identifying driver drowsiness. This study proposes a straightforward, cost-effective method for detecting driver drowsiness using a single channel. The contribution of this research is the discovery of drowsiness using discriminated features [moments features (M1, M2, M3, M4), roughness features (R1, R2, R3, R4), zero crossing rate (ZCR), sample entropy (SE) and median absolute deviation (MAD)] from publicly available datasets. A novel model was introduced in this study, which involved the fusion of wavelet transform Daubechies order 4 (WTDB4) and residue decomposition (RD) techniques for feature extraction. Various classification algorithms, including the least-square support vector machine (LSSVM) and ensemble models were compared in terms of their performance metrics. The algorithm that exhibited superior accuracy with reduced computational time was chosen to classify the driver's status into two groups: awake and drowsy. Notably, the proposed model achieved an impressive accuracy of 97.95%.
Developing a method to detect driver drowsiness based on a single EEG channel and discriminated features
Raed Mohammed Hussein; Loay E. George; Firas Sabar Miften
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 29 - 40
Driver drowsiness is one of the leading causes of road deaths and transportation industry dangers. Due to its direct evaluation of neurophysiological brain activity, electroencephalography (EEG) has been regarded as one of the most reliable physiological indicators for identifying driver drowsiness. This study proposes a straightforward, cost-effective method for detecting driver drowsiness using a single channel. The contribution of this research is the discovery of drowsiness using discriminated features [moments features (M1, M2, M3, M4), roughness features (R1, R2, R3, R4), zero crossing rate (ZCR), sample entropy (SE) and median absolute deviation (MAD)] from publicly available datasets. A novel model was introduced in this study, which involved the fusion of wavelet transform Daubechies order 4 (WTDB4) and residue decomposition (RD) techniques for feature extraction. Various classification algorithms, including the least-square support vector machine (LSSVM) and ensemble models were compared in terms of their performance metrics. The algorithm that exhibited superior accuracy with reduced computational time was chosen to classify the driver's status into two groups: awake and drowsy. Notably, the proposed model achieved an impressive accuracy of 97.95%.]]>
10.1504/IJISTA.2024.136521
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 29 - 40
Raed Mohammed Hussein
Loay E. George
Firas Sabar Miften
Iraqi Commission for Computers and Informatics, Informatics Institute of Postgraduate Studies, Baghdad, Iraq ' Computer Sciences Department, University of Information Technology and Communication, Baghdad, Iraq ' Computer Sciences Department, College of Education for Pure Science, University of Thi-Qar, Thi-Qar, Iraq
drowsy driving detection
electroencephalography
least square support vector machine
residue decomposition
wavelet transform
2024-02-05T23:20:50-05:00
Copyright © 2024 Inderscience Enterprises Ltd.
22
1
29
40
2024-02-05T23:20:50-05:00
-
Boosting speech recognition performance: a robust and accurate ensemble method based on HMMs
http://www.inderscience.com/link.php?id=136523
In this paper, we propose an ensemble method based on hidden Markov models (HMMs) for speech recognition. Our objective is to reduce the impact of the initial setting of training parameters on the final model while improving accuracy and robustness, particularly in speaker independent systems. The main idea is to exploit the sensitivity of HMMs to the initial setting of training parameters, thus creating diversity among the ensemble members. Additionally, we perform an experimental study to investigate the potential relationship between initial training parameters and ten diversity measures from literature. The proposed method is assessed on a standard dataset from the UCI machine-learning repository. Results demonstrate its effectiveness in terms of accuracy and robustness to intra-class variability, surpassing basic classifiers (HMM, KNN, NN, SVM) and some previous works in the literature including those using deep learning algorithms such as convolutional neural networks (CNNs) and long short-term memory (LSTM).
Boosting speech recognition performance: a robust and accurate ensemble method based on HMMs
Samira Hazmoune; Fateh Bougamouza; Smaine Mazouzi; Mohamed Benmohammed
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 41 - 76
In this paper, we propose an ensemble method based on hidden Markov models (HMMs) for speech recognition. Our objective is to reduce the impact of the initial setting of training parameters on the final model while improving accuracy and robustness, particularly in speaker independent systems. The main idea is to exploit the sensitivity of HMMs to the initial setting of training parameters, thus creating diversity among the ensemble members. Additionally, we perform an experimental study to investigate the potential relationship between initial training parameters and ten diversity measures from literature. The proposed method is assessed on a standard dataset from the UCI machine-learning repository. Results demonstrate its effectiveness in terms of accuracy and robustness to intra-class variability, surpassing basic classifiers (HMM, KNN, NN, SVM) and some previous works in the literature including those using deep learning algorithms such as convolutional neural networks (CNNs) and long short-term memory (LSTM).]]>
10.1504/IJISTA.2024.136523
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 41 - 76
Samira Hazmoune
Fateh Bougamouza
Smaine Mazouzi
Mohamed Benmohammed
Department of Computer Science, Faculty of Sciences, University of 20 Août 1955-Skikda, Skikda, Algeria ' Department of Computer Science, Faculty of Sciences, University of 20 Août 1955-Skikda, Skikda, Algeria ' Department of Computer Science, Faculty of Sciences, University of 20 Août 1955-Skikda, Skikda, Algeria ' Department of Software Technologies and Information Systems, Faculty of New Technologies of Information and Communication, University Constantine 2, Constantine, Algeria
speech recognition
inter-speaker variability
robustness
accuracy
HMM
multiple modelling
ensemble methods
diversity
2024-02-05T23:20:50-05:00
Copyright © 2024 Inderscience Enterprises Ltd.
22
1
41
76
2024-02-05T23:20:50-05:00
-
Design and development of political rider competitive swarm optimiser enabled deep learning model for air quality detection
http://www.inderscience.com/link.php?id=136526
The air quality prediction process is a more significant one for air pollution prevention and management because air pollution becomes crueller. The precise identification of air quality has become a more significant concern for controlling air pollution. Recently, the weight of particulate matter (PM) on the human physical condition has become an important research area. In this paper, the political rider competitive swarm optimiser (PRCSO)-based deep recurrent neural network (DRNN) algorithm is devised for air quality and carbon monoxide prediction. The missing value imputation scheme is employed to perform pre-processing. Moreover, technical indicators and location information are extracted for the prediction process. The DRNN is employed for prediction, which is trained by the PRCSO and the training process is performed based on every location independently. The PRCSO-based DRNN outperforms existing techniques in terms of mean square error (MSE) of 0.0313, and mean absolute percentage error (MAPE) of 3.08%.
Design and development of political rider competitive swarm optimiser enabled deep learning model for air quality detection
Deepika Dadasaheb Patil; T.C. Thanuja; Bhuvaneshwari C. Melinamath
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 77 - 104
The air quality prediction process is a more significant one for air pollution prevention and management because air pollution becomes crueller. The precise identification of air quality has become a more significant concern for controlling air pollution. Recently, the weight of particulate matter (PM) on the human physical condition has become an important research area. In this paper, the political rider competitive swarm optimiser (PRCSO)-based deep recurrent neural network (DRNN) algorithm is devised for air quality and carbon monoxide prediction. The missing value imputation scheme is employed to perform pre-processing. Moreover, technical indicators and location information are extracted for the prediction process. The DRNN is employed for prediction, which is trained by the PRCSO and the training process is performed based on every location independently. The PRCSO-based DRNN outperforms existing techniques in terms of mean square error (MSE) of 0.0313, and mean absolute percentage error (MAPE) of 3.08%.]]>
10.1504/IJISTA.2024.136526
International Journal of Intelligent Systems Technologies and Applications, Vol. 22, No. 1 (2024) pp. 77 - 104
Deepika Dadasaheb Patil
T.C. Thanuja
Bhuvaneshwari C. Melinamath
Department of Computer Science and Engineering, Vishveshvaraya Technological University, Belgavi, India; Sanjay Ghodawat University, Kolhapur, India ' Department of VLSI and Embedded System, Visvesvaraya Technological University, Belagavi, India ' Department of Computer Science and Engineering, BMSIT&M College, Bangalore, India
air quality prediction
carbon monoxide prediction
deep recurrent neural network
DRNN
political optimiser
relative strength index
mean square error
MSE
mean absolute percentage error
MAPE
2024-02-05T23:20:50-05:00
Copyright © 2024 Inderscience Enterprises Ltd.
22
1
77
104
2024-02-05T23:20:50-05:00