Mardin Meslek Yüksekokulu

Permanent URI for this communityhttps://hdl.handle.net/20.500.12514/28

Browse

Now showing 1 - 20 of 29

Balina Optimizasyon Algoritması Kullanılarak Türkiye’nin Uzun Vadeli Enerji Tüketimi Tahmini
(IEEE Xplore, 2021) Babaoğlu, Merve; Haznedar, Bülent
Enerji, ülkelerin sürdürülebilir kalkınmaları için en önemli konu başlıklarından biridir. Kullanılan enerjinin tükenebilir olması, birçok enerji kaynağını ithal ediyor olması ve çevresel faktörlerden dolayı Türkiye için gelecekte enerji ihtiyacının ne kadar olabileceğinin tahmin edilebilmesi büyük önem taşımaktadır. Bu çalışmada Türkiye’nin 2040 yılına kadarki enerji tüketim tahminini yapabilmek adına, sezgisel algoritmalardan balina optimizasyon algoritması (BOA) tercih edilmiştir. Balina optimizasyon algoritmasının performansını belirleyebilmek için elde edilen veriler, genetik algoritma (GA) sonuçları ile karşılaştırılmıştır. Tüm modeller doğrusal olarak düzenlenip sonuç alınmıştır. Enerji talebini etkileyen gayri safi yurtiçi hasıla (GSYH), nüfus, ithalat ve ihracat gibi bağımsız değişkenlerin 1990-2019 yılları arasındaki verileri kullanılmıştır. Sonuçların doğruluğunu hesaplayabilmek için geçmiş 30 yılın modellenmesi sağlanmıştır. En uygun model elde edildikten sonra gelecek 20 yıl için 4 farklı senaryoya göre tahminler yapılmıştır.
A Class Activation Map-Based Interpretable Transfer Learning Model for Automated Detection of ADHD from fMRI Data
(Sage Journals, 2022) Uyulan, Caglar; Erguzel, Turker Tekin; Türk, Ömer; Farhad, Shams; Metin, Bariş; Tarhan, Nevzat; Türk, Ömer
Automatic detection of Attention Deficit Hyperactivity Disorder (ADHD) based on the functional Magnetic Resonance Imaging (fMRI) through Deep Learning (DL) is becoming a quite useful methodology due to the curse of-dimensionality problem of the data is solved. Also, this method proposes an invasive and robust solution to the variances in data acquisition and class distribution imbalances. In this paper, a transfer learning approach, specifically ResNet-50 type pre-trained 2D-Convolutional Neural Network (CNN) was used to automatically classify ADHD and healthy children. The results demonstrated that ResNet-50 architecture with 10-k cross-validation (CV) achieves an overall classification accuracy of 93.45%. The interpretation of the results was done via the Class Activation Map (CAM) analysis which showed that children with ADHD differed from controls in a wide range of brain areas including frontal, parietal and temporal lobes.
Classification and analysis of epileptic EEG recordings using convolutional neural network and class activation mapping
(2021) Yildiz, Abdulnasir; Zan, Hasan; Said, Sherif; Zan, Hasan
Electrical bio-signals have the potential to be used in different applications due to their hidden nature and their ability to facilitate liveness detection. This paper investigates the feasibility of using the Convolutional Neural Network (CNN) to classify and analyze electroencephalogram (EEG) data with their time-frequency representations and class activation mapping (CAM) to detect epilepsy disease. Several types of pre-trained CNNs are employed for a multi-class classification task (AlexNet, GoogLeNet, ResNet-18, and ResNet-50) and their results are compared. Also, a novel convolutional neural network architecture comprised of two horizontally concatenated GoogLeNets is proposed with two inputs scalograms and spectrogram of the eplictic EEG signal. Four segment lengths (4097, 2048, 1024, and 512 sampling points) with three time-frequency representations (short-time Fourier, Wavelet, and Hilbert-Huang transform) are statistically evaluated. The dataset used in this research is collected at the University of Bonn. The dataset is reorganized as normal, interictal, and ictal. The maximum achieved accuracies for 4097, 2048, 1024, and 512 sampling points are 100 %, 100 %, 100 %, and 99.5 % respectively. The CAM method is used to analyze discriminative regions of time-frequency representations of EEG segments and networks' decisions. This method showed CNN models used different time and frequency regions of input images for each class with correct and incorrect predictions.
Classification and analysis of epileptic EEG recordings using convolutional neural network and class activation mapping
(Biomedical Signal Processing and Control, 2021) Zan, Hasan; Yıldız, Abdulnasir; Said, Sherif; Zan, Hasan
Electrical bio-signals have the potential to be used in different applications due to their hidden nature and their ability to facilitate liveness detection. This paper investigates the feasibility of using the Convolutional Neural Network (CNN) to classify and analyze electroencephalogram (EEG) data with their time-frequency representations and class activation mapping (CAM) to detect epilepsy disease. Several types of pre-trained CNNs are employed for a multi-class classification task (AlexNet, GoogLeNet, ResNet-18, and ResNet-50) and their results are compared. Also, a novel convolutional neural network architecture comprised of two horizontally concatenated GoogLeNets is proposed with two inputs scalograms and spectrogram of the eplictic EEG signal. Four segment lengths (4097, 2048, 1024, and 512 sampling points) with three time-frequency representations (short-time Fourier, Wavelet, and Hilbert-Huang transform) are statistically evaluated. The dataset used in this research is collected at the University of Bonn. The dataset is reorganized as normal, interictal, and ictal. The maximum achieved accuracies for 4097, 2048, 1024, and 512 sampling points are 100 %, 100 %, 100 %, and 99.5 % respectively. The CAM method is used to analyze discriminative regions of time-frequency representations of EEG segments and networks' decisions. This method showed CNN models used different time and frequency regions of input images for each class with correct and incorrect predictions.
Classification of electroencephalogram records related to cursor movements with a hybrid method based on deep learning
(Wiley Online Library, 2021) Türk, Ömer; Türk, Ömer
In brain computer interface (BCI), many transformation methods are used whenprocessing electroencephalogram (EEG) signals. Thus, the EEG can be represen-ted in different domains. However, designing an EEG-based BCI system withoutany transformation technique is a challenge. For this purpose, in this study, aBCI model is proposed without any transformation. The classification of cursordown and cursor up movements using the EEG signals received from the brain isaimed at in the proposed model. The EEG patterns were classified using twomethods. Firstly, EEG signals were classified by classic convolutional neural net-work (CNN). Secondly, proposed hybrid structure obtained the EEG features,which were classified by k-NN and SVM, using CNN. Classification with CNNarchitecture gave a result of 68.15% while the hybrid method using k-NN andSVM classifiers yielded 97.55% and 97.61% respectively. The hybrid proposedmethod were more successful than the studies in the literature.
A comprehensive review on detection of cyber-attacks: Data sets, methods, challenges, and future research directions
(ScienceDirect, 2022) Ahmetoglu, Huseyin; Das, Resul
Rapid developments in network technologies and the amount and scope of data transferred on networks are increasing day by day. Depending on this situation, the density and complexity of cyber threats and attacks are also expanding. The ever-increasing network density makes it difficult for cyber-security professionals to monitor every movement on the network. More frequent and complex cyber-attacks make the detection and identification of anomalies in network events more complex. Machine learning offers various tools and techniques for automating the detection of cyber attacks and for rapid prediction and analysis of attack types. This study discusses the approaches to machine learning methods used to detect attacks. We examined the detection, classification, clustering, and analysis of anomalies in network traffic. We gave the cyber-security focus, machine learning methods, and data sets used in each study we examined. We investigated which feature selection or dimension reduction method was applied to the data sets used in the studies. We presented in detail the types of classification carried out in these studies, which methods were compared with other methods, the performance metrics used, and the results obtained in tables. We examined the data sets of network attacks presented as open access. We suggested a basic taxonomy for cyber attacks. Finally, we discussed the difficulties encountered in machine learning applications used in network attacks and their solutions.
The convolutional neural network approach from electroencephalogram signals in emotional detection
(Concurrency Computation, 2021) Türk, Ömer; Özerdem, Mehmet Siraç; Türk, Ömer
Although brain-computer interfaces (BCI) progress rapidly, the desired success has not been achieved yet. One of these BCI is to detect emotional states in humans. An emotional state is a brain activity consisting of hormonal and mental reasons in the face of events. Emotions can be detected by electroencephalogram (EEG) signals due to these activities. Being able to detect the emotional state from EEG signals is important in terms of both time and cost. In this study, a method is proposed for the detection of the emotional state by using EEG signals. In the proposed method, we aim to classify EEG signals without any transform (Fourier transform, wavelet transform, etc.) or feature extraction method as a pre-processing. For this purpose, convolutional neural networks (CNNs) are used as classifiers, together with SEED EEG dataset containing three different emotional (positive, negative, and neutral) states. The records used in the study were taken from 15 participants in three sessions. In the proposed method, raw channel-time EEG recordings are converted into 28 × 28 size pattern segments without pre-processing. The obtained patterns are then classified in the CNN. As a result of the classification, three emotion performance averages of all participants are found to be 88.84%. Based on the participants, the highest classification performance is 93.91%, while the lowest classification performance is 77.70%. Also, the average f-score is found to be 0.88 for positive emotion, 0.87 for negative emotion, and 0.89 for neutral emotion. Likewise, the average kappa value is 0.82 for positive emotion, 0.81 for negative emotion, and 0.83 for neutral emotion. The results of the method proposed in the study are compared with the results of similar studies in the literature. We conclude that the proposed method has an acceptable level of performance.
Developed analytical expression for current harmonic distortion of the PV system’s inverter in relation to the solar irradiance and temperature
(Electrical Engineering, 2021) Adak, Süleyman; Cangi Hasan; Eid Bilal,; Yılmaz Ahmet Sedar
This paper deals with modeling and simulation of the total harmonic distortion of the current (THDI) dispatched from the inverter and connected to nonlinear load. The change of THDI was examined in relation to the ambient temperature (T) and solar irradiance (G). The developed model is being used to extract parameters for a given THDI as a function of temperature and solar radiation. This study outlines the working principle of photovoltaic (PV) panel as well as PV array. Off-grid PV system is modeled by using Matlab/Simulink program, and detailed analytical study has been carried out in this work. The design, modeling and simulation of this study are performed from 50 up to 988 W/m2 for solar irradiance. Harmonic components have negative effects on the steady-voltage stability of the PV system. Therefore, analytical expression is needed for steady-state stability analysis in order to reduce negative effects. Hence, two analytical expressions of THDI were obtained by two new different methods which are statistical package for the social sciences program and genetic expression programming. Eventually, two different methods have been verified by the Matlab/Simulink program in order to find out THDI and demonstrated the effectiveness of the proposed strategy. As a result of this study, it is observed that input current THDI of nonlinear load is too high at low irradiance. It is suggested that active harmonic filters should be used at low irradiance in order to produce better quality energy and avoid damages in the PV system.
Energy and exergy analysis of an industrial corn dryer operated by two different fuels
(International Journal of Exergy, 2021) Ünal, Fatih
In this study, the data obtained after converting an industrial horizontal type corn dryer that meets its drying air temperature from coal to natural gas was compared by thermodynamic analyses. Before starting the drying process, it was assumed that the corn type DKC6050 with 24-25% corn inlet humidity dries when it reaches approximately 14% corn outlet humidity, which is the storage condition after the drying process. Energy and exergy efficiencies, drying rates, unit drying costs, specific moisture extraction rate, and specific energy consumption values of the analysed systems were determined using the data obtained from the experiments carried out at 90, 100 and 110 C drying temperatures. On the other hand, it was also determined that the unit drying cost was approximately 0.1-0.45 €/kg and the specific energy consumption was less than approximately 1,000-8,000 kJ/kgwater. Also, emission values released to the environment were calculated for both systems based on the amount of energy required for drying.
Energy, exergy and exergoeconomic analysis of solar-assisted vertical ground source heat pump system for heating season
(KOREAN SOC MECHANICAL ENGINEERS, 2018) Unal, Fatih; Temir, Galip; Koten, Hasan
The purpose of this study is to evaluate the experimental performance of a solar assisted vertical ground source heat pump system (VGSHP) for the winter climatic conditions of Mardin, which is in the South-Eastern Anatolia region of Turkey. For this aim, an experimental analysis was performed on solar assisted VGSHP system, which was designed to meet the heating needs of an experimental room, during the heating season (10.01.2013/03.31.2014). The experimentally obtained results were used to calculate energy, exergy and exergoeconomic analyses of the system and its components. The energy efficiency, exergy efficiency and exergoeconomic factors of the entire system were 67.36 %, 27.40 % and 60.51 %, respectively. In this study, the system was proposed for disseminating the use of alternative technologies supported by renewable energy systems and it has been tested for the first time in Mardin to meet its heating needs with convectional systems. The experimental results showed that the proposed solar assisted VGSHP system can be used for residential heating in Mardin and similar regions. As a result, it has been detected that the system is very effective in both reducing energy consumption and decreasing emissions of green-house gases.
FPGA simulation of chaotic tent map-based S-Box design
(Wiley Online Library, 2022) Türk, Ömer; Türk, Ömer
The chaotic system has a characteristically random behavior by nature, and these systems have their own characteristics in a completely deterministic structure. This feature of a chaotic system makes it difficult to predict encryptions designed based on such a system. Thanks to this unpredictable and strong feature, maps produced from chaotic systems are an important alternative in the field of encryption. One of the structures obtained by employing chaotic maps is the substitution box. S-Box, which provides the confusion principle used in block ciphers, is the main block that dynamically replaces unencrypted data with confidential data and makes a significant contribution to ensuring high security in the encryption system. Therefore, S-Boxes hold a critical role in block ciphers. Speed and reliability are important parameters in the creation of this main block. Especially, applications performed on hardware are more reliable and high performance. Therefore, in this study, an S-Box was designed using fieldprogrammable gate arrays (FPGA) simulation from a chaotic tent map to create a fast and reliable S-Box because FPGAs offer solutions that may be important in this field considering their fast and customizable architecture. In the proposed method, the S-Box was created in 0.16 s. In addition, the dynamic properties of the chaotic tent map were analyzed with Lyapunov exponents, and the NIST SP 800-22 test was applied for the information encryption suitability of the proposed chaotic system. Also, to test the reliability of the produced S-Box structures, SAC, non-linearity, bit independence criteria, and input/output XOR distribution table metrics were implemented. The results showed that the proposed chaotic map was dynamic and passed the reliability tests successfully.
Harmonics Mitigation of Stand‑Alone Photovoltaic System Using LC Passive Filter
(Journal of Electrical Engineering & Technology, 2021) Adak, Süleyman
This article investigates modeling and simulation of the of-grid photovoltaic (PV) system, and elimination of harmonic components using an LC passive flter. Pulse width modulation (PWM) inverter is used to convert the direct current to alternating current. It is very important in terms of energy quality that the inverter output current total harmonic distortion (THDI) is below the value given by standards. Harmonic components have negatively efect on of-grid PV power system. THDI should be kept below a certain level in order to prevent damage to the equipment in the of-grid system and to ensure a higher quality energy fow to reduce the total harmonic distortion (THD) of the solar inverter output current; LC passive flter must be connected to the output of the PWM inverter. There are many types of passive flters for solar inverters. One of the most widely used flter types is the LC flter. LC flters are used in of-grid systems. LC flter is smaller in size and lower cost than other flters. But it is more complicated to determine the parameters of the LC flter. Therefore, in order for the system to remain in a steady state, the parameters must be accurately calculated and analyzed. In this study, the output power of the solar inverter, switching frequency, bus voltage etc. values were determined and LC flter parameters were calculated. Since high inductance values are used in LC flters, the voltage drop increases in these flters. To reduce the voltage drop, the DC bus voltage must be increased, which increases the switching losses. LC flter is connected between the inverter and the nonlinear load to flter the harmonic components produced by the DC/DC boost converter, DC/AC inverter and non-linear load. Matlab/Simulink program was used in Simulation and analysis of of-grid solar system. Solar inverter output current THD was measured as 91.55%. After the LC flter is connected to the system, this value has dropped to 2.62%.
Harmonics Mitigation of Stand-Alone Photovoltaic System Using LC Passive Filter
(Journal of Electrical Engineering and Technology, 2021) Adak, Süleyman
This article investigates modeling and simulation of the off-grid photovoltaic (PV) system, and elimination of harmonic components using an LC passive filter. Pulse width modulation (PWM) inverter is used to convert the direct current to alternating current. It is very important in terms of energy quality that the inverter output current total harmonic distortion (THDI) is below the value given by standards. Harmonic components have negatively effect on off-grid PV power system. THDI should be kept below a certain level in order to prevent damage to the equipment in the off-grid system and to ensure a higher quality energy flow to reduce the total harmonic distortion (THD) of the solar inverter output current; LC passive filter must be connected to the output of the PWM inverter. There are many types of passive filters for solar inverters. One of the most widely used filter types is the LC filter. LC filters are used in off-grid systems. LC filter is smaller in size and lower cost than other filters. But it is more complicated to determine the parameters of the LC filter. Therefore, in order for the system to remain in a steady state, the parameters must be accurately calculated and analyzed. In this study, the output power of the solar inverter, switching frequency, bus voltage etc. values were determined and LC filter parameters were calculated. Since high inductance values are used in LC filters, the voltage drop increases in these filters. To reduce the voltage drop, the DC bus voltage must be increased, which increases the switching losses. LC filter is connected between the inverter and the nonlinear load to filter the harmonic components produced by the DC/DC boost converter, DC/AC inverter and non-linear load. Matlab/Simulink program was used in Simulation and analysis of off-grid solar system. Solar inverter output current THD was measured as 91.55%. After the LC filter is connected to the system, this value has dropped to 2.62%.
Identification of cotton and corn plant areas by employing deep transformer encoder approach and different time series satellite images: A case study in Diyarbakir, Turkey
(ScienceDirect, 2023) Türk, Ömer; Şimşek Bağcı, Reyhan; Acar, Emrullah; Türk, Ömer
It is very important to determine the crops in the agricultural field in a short time and accurately. Thanks to the satellite images obtained from remote sensing sensors, information can be obtained on many subjects such as the detection and development of agricultural products and annual product forecasting. In this study, it is aimed to automatically detect agricultural crops (corn and cotton) by using Sentinel-1 and Landsat-8 satellite image indexes via a new deep learning approach (Deep Transformer Encoder). This work was carried out in several stages, respectively. In the first stage, a pilot area was determined to obtain Sentinel-1 and Landsat-8 satellite images of agricultural crops used in this study. In the second stage, the coordinates of 100 sample points from this pilot area were taken with the help of GPS and these coordinates were then transferred to Sentinel-1 and Landsat-8 satellite images. In the next step, reflection and backscattering values were obtained from the pixels of the satellite images corresponding to the sample points of these agricultural crops. While creating the data sets of satellite images, the months of June, July, August and September for the years 2016–2021, when the development and harvesting times of agricultural products are close to each other, were preferred. The image data set used in the study consists of a total of 434 images for Sentinel-1 satellite and a total of 693 images for Landsat-8. At the last stage, the datasets obtained from different satellite images were evaluated in three different categories for crop identification with the aid of Deep Transformer Encoder approach. These are: (1-) Crop identification with only Sentinel-1 dataset, (2-) Crop identification only with Landsat-8 dataset, (3-) Crop identification with both Sentinel-1 and Landsat-8 datasets. The results showed that 85%, 95% and 87.5% accuracy values were obtained from the band parameters of Sentinel-1 dataset, Landsat-8 dataset and Sentinel-1&Landsat-8 datasets, respectively
Intangible cultural heritage: Local food and tourism
(Cultural Heritage and Tourism, 2021) Bozkurt, İrem; Yıldırım, Enes
Intangible cultural heritage: Local food and tourism
The Karapapaks and their shifting loyalties on the imperial borderlands during the nineteenth century
(Taylor & Francis Online, 2022) Çiftçi, Erdal
The Karapapaks were one of the less known native Turkish ethnic groups of the Transcaucasia, who overwhelmingly took refuge in the Ottoman and Qajar Empires in the late 1820s, after the expansion of Tsarist Russia into their homelands. This paper analyses how the literature regarding Karapapak movements and society was overwhelmingly shaped by selective, essentialist, and anachronistic approaches by some historians in Turkey and Iran. While the former determined that they were a loyal pro-Ottoman and pro-Sunni Karapapak society, the latter constructed an opposing pro-Iranian and pro-Shiite narrative. This paper deconstructs both approaches, and asserts that the collective ethnic and sectarian identities of this society played a secondary role in regards to influencing their cross-border movements. This paper argues that the approach of the current literature cannot explain this borderland society’s perpetual, multiple and multi-directional cross-border movement. Instead, the Karapapaks often manoeuvred the frontiers of the empires, and defected to another empire when it was necessary to, first and foremost, satisfy the needs of their own society, over those of any imperial allies.
Local Pattern Transformation-Based convolutional neural network for sleep stage scoring
(2023) Zan, Hasan; Yıldız, Abdulnasır; Zan, Hasan
Sleep stage scoring is essential for the diagnosis and treatment of sleep disorders. However, manual sleep scoring is a tedious, time-consuming, and subjective task. Therefore, this paper proposes a novel framework based on local pattern transformation (LPT) methods and convolutional neural networks for automatic sleep stage scoring. Unlike in previous works in other fields, these methods were not employed for manual feature extraction, which requires expert knowledge and the pipeline behind it might bias results. The transformed signals were directly fed into a CNN model (called EpochNet) that can accept multiple successive epochs. The model learns features from multiple input epochs and considers inter-epoch context during classification. To evaluate and validate the effectiveness of the proposed approach, we conducted several experiments on the Sleep-EDF dataset. Four LPT methods, including One-dimensional Local Binary Pattern (1D-LBP), Local Neighbor Descriptive Pattern (LNDP), Local Gradient Pattern (LGP), and Local Neighbor Gradient Pattern (LNGP), and different polysomnography (PSG) signals were analyzed as sequence length (number of input epochs) increased from one to five. 1D-LBP and LNDP achieved similar performances, outperforming other LPT methods that are less sensitive to local variations. The best performance was achieved when an input sequence containing five epochs of PSG signals transformed by 1D-LBP was employed. The best accuracy, F1 score, and Kohen's kappa coefficient were 0.848, 0.782, and 0.790, respectively. The results showed that our approach can achieve comparable performance to other state-of-the-art methods while occupying fewer computing resources because of the compact size of EpochNet.
Local Pattern Transformation-Based convolutional neural network for sleep stage scoring
(ScienceDirect, 2023) Zan, Hasan; Yildiz, Abdulnasır; Zan, Hasan
Sleep stage scoring is essential for the diagnosis and treatment of sleep disorders. However, manual sleep scoring is a tedious, time-consuming, and subjective task. Therefore, this paper proposes a novel framework based on local pattern transformation (LPT) methods and convolutional neural networks for automatic sleep stage scoring. Unlike in previous works in other fields, these methods were not employed for manual feature extraction, which requires expert knowledge and the pipeline behind it might bias results. The transformed signals were directly fed into a CNN model (called EpochNet) that can accept multiple successive epochs. The model learns features from multiple input epochs and considers inter-epoch context during classification. To evaluate and validate the effectiveness of the proposed approach, we conducted several experiments on the Sleep-EDF dataset. Four LPT methods, including One-dimensional Local Binary Pattern (1D-LBP), Local Neighbor Descriptive Pattern (LNDP), Local Gradient Pattern (LGP), and Local Neighbor Gradient Pattern (LNGP), and different polysomnography (PSG) signals were analyzed as sequence length (number of input epochs) increased from one to five. 1D-LBP and LNDP achieved similar performances, outperforming other LPT methods that are less sensitive to local variations. The best performance was achieved when an input sequence containing five epochs of PSG signals transformed by 1D-LBP was employed. The best accuracy, F1 score, and Kohen’s kappa coefficient were 0.848, 0.782, and 0.790, respectively. The results showed that our approach can achieve comparable performance to other state-ofthe-art methods while occupying fewer computing resources because of the compact size of EpochNet.
Multi-task learning for arousal and sleep stage detection using fully convolutional networks
(IOP Publishing, 2023) Zan, Hasan; Yıldız, Abdulnasir; Zan, Hasan
Objective: Sleep is a critical physiological process that plays a vital role in maintaining physical and mental health. Accurate detection of arousals and sleep stages is essential for the diagnosis of sleep disorders, as frequent and excessive occurrences of arousals disrupt sleep stage patterns and lead to poor sleep quality, negatively impacting physical and mental health. Polysomnography is a traditional method for arousal and sleep stage detection that is time-consuming and prone to high variability among experts. Approach: In this paper, we propose a novel multi-task learning approach for arousal and sleep stage detection using fully convolutional neural networks. Our model, FullSleepNet, accepts a full-night single-channel EEG signal as input and produces segmentation masks for arousal and sleep stage labels. FullSleepNet comprises four modules: a convolutional module to extract local features, a recurrent module to capture long-range dependencies, an attention mechanism to focus on relevant parts of the input, and a segmentation module to output final predictions. Main results: By unifying the two interrelated tasks as segmentation problems and employing a multi-task learning approach, FullSleepNet achieves state-of-the-art performance for arousal detection with an area under the precision-recall curve of 0.70 on Sleep Heart Health Study and Multi-Ethnic Study of Atherosclerosis datasets. For sleep stage classification, FullSleepNet obtains comparable performance on both datasets, achieving an accuracy of 0.88 and an F1-score of 0.80 on the former and an accuracy of 0.83 and an F1-score of 0.76 on the latter. Significance: Our results demonstrate that FullSleepNet offers improved practicality, efficiency, and accuracy for the detection of arousal and classification of sleep stages using raw EEG signals as input.
Multi-task learning for arousal and sleep stage detection using fully convolutional networks
(2023) Zan, Hasan; Yıldız, Abdulnasır; Zan, Hasan
Objective. Sleep is a critical physiological process that plays a vital role in maintaining physical and mental health. Accurate detection of arousals and sleep stages is essential for the diagnosis of sleep disorders, as frequent and excessive occurrences of arousals disrupt sleep stage patterns and lead to poor sleep quality, negatively impacting physical and mental health. Polysomnography is a traditional method for arousal and sleep stage detection that is time-consuming and prone to high variability among experts. Approach. In this paper, we propose a novel multi-task learning approach for arousal and sleep stage detection using fully convolutional neural networks. Our model, FullSleepNet, accepts a full-night single-channel EEG signal as input and produces segmentation masks for arousal and sleep stage labels. FullSleepNet comprises four modules: a convolutional module to extract local features, a recurrent module to capture long-range dependencies, an attention mechanism to focus on relevant parts of the input, and a segmentation module to output final predictions. Main results. By unifying the two interrelated tasks as segmentation problems and employing a multi-task learning approach, FullSleepNet achieves state-of-the-art performance for arousal detection with an area under the precision-recall curve of 0.70 on Sleep Heart Health Study and Multi-Ethnic Study of Atherosclerosis datasets. For sleep stage classification, FullSleepNet obtains comparable performance on both datasets, achieving an accuracy of 0.88 and an F1-score of 0.80 on the former and an accuracy of 0.83 and an F1-score of 0.76 on the latter. Significance. Our results demonstrate that FullSleepNet offers improved practicality, efficiency, and accuracy for the detection of arousal and classification of sleep stages using raw EEG signals as input.

Browse

Browsing Mardin Meslek Yüksekokulu by Access Right "info:eu-repo/semantics/closedAccess"