: In this paper, lightweight deep learning methods are proposed to recognize multichannel electromyography (EMG) signals against varying contraction levels. The classical machine learning, and signal processing methods namely, linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), root mean square (RMS), and waveform length (WL) are adopted to convolutional neural network (CNN), and long short-term memory neural network (LSTM). Eight-channel recordings of nine amputees from a publicly available dataset are used for training and testing the proposed models considering prosthetic control strategies. Six class hand movements with three contraction levels are applied to WL and RMS-based feature extraction. After that, they are formed into appropriate input dimensions, and classified using the LDA, QDA, LDA-CNN, QDA-CNN, LSTM, and CNN. Depending on three prosthetic EMG validation approaches (Scheme 1-3), the accuracy rates of 41.68%, and 47.27% are yielded by LDA, and QDA with 32-dimensional RMS, and WL features while CNN with 2 × 16 input has 82.87% (up to 88.10%). The effect of the learnable filters of the DL approaches, and signal windowing on the success rate and delay time are discussed in the paper. The simulations show that 2D-CNN (accuracy of 82.87% with 1.7 ms delay) can be successfully adapted to prosthetic control devices.
{"title":"Lightweight deep neural network models for electromyography signal recognition for prosthetic control","authors":"A. Mert","doi":"10.55730/1300-0632.4012","DOIUrl":"https://doi.org/10.55730/1300-0632.4012","url":null,"abstract":": In this paper, lightweight deep learning methods are proposed to recognize multichannel electromyography (EMG) signals against varying contraction levels. The classical machine learning, and signal processing methods namely, linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), root mean square (RMS), and waveform length (WL) are adopted to convolutional neural network (CNN), and long short-term memory neural network (LSTM). Eight-channel recordings of nine amputees from a publicly available dataset are used for training and testing the proposed models considering prosthetic control strategies. Six class hand movements with three contraction levels are applied to WL and RMS-based feature extraction. After that, they are formed into appropriate input dimensions, and classified using the LDA, QDA, LDA-CNN, QDA-CNN, LSTM, and CNN. Depending on three prosthetic EMG validation approaches (Scheme 1-3), the accuracy rates of 41.68%, and 47.27% are yielded by LDA, and QDA with 32-dimensional RMS, and WL features while CNN with 2 × 16 input has 82.87% (up to 88.10%). The effect of the learnable filters of the DL approaches, and signal windowing on the success rate and delay time are discussed in the paper. The simulations show that 2D-CNN (accuracy of 82.87% with 1.7 ms delay) can be successfully adapted to prosthetic control devices.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85164036","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The diagnosis of diabetes, a prevalent global health condition, is crucial for preventing severe complications. In recent years, there has been a growing effort to develop intelligent diagnostic systems for diabetes utilizing machine learning (ML) algorithms. Despite these efforts, achieving high accuracy rates using such systems remains a significant challenge. Recent advancements in ensemble ML methods offer promising opportunities for early detection of diabetes, as they are known to be faster and more cost-effective than traditional approaches. Therefore, this study proposes a practical framework for diagnosing diabetes that involves three stages. The data preprocessing stage encompasses several crucial tasks, including handling missing values, identifying outliers, balancing the data, normalizing the data, and selecting relevant features. Subsequently, the hyperparameters of the ML algorithms are fine-tuned using grid search to improve their performance. In the final stage, the framework employs ensemble techniques such as bagging, boosting, and stacking to combine multiple ML algorithms and further enhance their predictive capability. Pima Indians Diabetes Database open-access dataset was used to test the performance of the proposed models. The experimental results of this framework indicate the superiority of ensemble methods in diagnosing diabetes compared to individual ML models. The stacking method achieved the best accuracy among the ensemble methods, with the stacked random forest (RF) and support vector machine (SVM) model attaining an accuracy of 97.50%. Among the bagging methods, the RF model yielded the highest accuracy, while among the boosting methods, eXtreme Gradient Boosting (XGB) model achieved the highest accuracy rates of 97.20% and 97.10%, respectively. Moreover, our proposed framework outperforms other ML models as confirmed by the comparison. The study has demonstrated that ensemble methods are crucial for accurate diabetes diagnosis, enabling early detection through efficient preprocessing and calibrated models.
{"title":"A practical framework for early detection of diabetes using ensemble machine learning models","authors":"Qusay Saihood, Emrullah Sonuç","doi":"10.55730/1300-0632.4013","DOIUrl":"https://doi.org/10.55730/1300-0632.4013","url":null,"abstract":"The diagnosis of diabetes, a prevalent global health condition, is crucial for preventing severe complications. In recent years, there has been a growing effort to develop intelligent diagnostic systems for diabetes utilizing machine learning (ML) algorithms. Despite these efforts, achieving high accuracy rates using such systems remains a significant challenge. Recent advancements in ensemble ML methods offer promising opportunities for early detection of diabetes, as they are known to be faster and more cost-effective than traditional approaches. Therefore, this study proposes a practical framework for diagnosing diabetes that involves three stages. The data preprocessing stage encompasses several crucial tasks, including handling missing values, identifying outliers, balancing the data, normalizing the data, and selecting relevant features. Subsequently, the hyperparameters of the ML algorithms are fine-tuned using grid search to improve their performance. In the final stage, the framework employs ensemble techniques such as bagging, boosting, and stacking to combine multiple ML algorithms and further enhance their predictive capability. Pima Indians Diabetes Database open-access dataset was used to test the performance of the proposed models. The experimental results of this framework indicate the superiority of ensemble methods in diagnosing diabetes compared to individual ML models. The stacking method achieved the best accuracy among the ensemble methods, with the stacked random forest (RF) and support vector machine (SVM) model attaining an accuracy of 97.50%. Among the bagging methods, the RF model yielded the highest accuracy, while among the boosting methods, eXtreme Gradient Boosting (XGB) model achieved the highest accuracy rates of 97.20% and 97.10%, respectively. Moreover, our proposed framework outperforms other ML models as confirmed by the comparison. The study has demonstrated that ensemble methods are crucial for accurate diabetes diagnosis, enabling early detection through efficient preprocessing and calibrated models.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79218572","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Elang Aji Defrianto, Atya Saniah, S. F. Rahman, N. R. Poespawati
Perovskite solar cells are the most cutting-edge photovoltaic technology having high efficiency and short fabrication time. In recent decades, there has been a significant rise in the study of the usage of carbon materials in perovskite solar cells because of low cost and earth abundance. Several studies have been conducted to mix hole transport materials with carbon materials to improve the hole extraction capability. Nevertheless, no research has reported using CuSCN on different carbon electrodes on perovskite solar cells. In this research, various carbon materials, including carbon nanotubes (CNT), graphite, activated carbon, and reduced graphene oxide (rGO), are mixed with CuSCN. The carbon materials and CuSCN were mixed by ball mill and then deposited using the doctor blading method to become an electrode layer. The existence of CuSCN in carbon materials was proved by conducting the energy dispersive X-ray test. CNT mixed with CuSCN material exhibits the highest electrical conductivity indicated by ID/IG ratio of 1.22 using Raman spectroscopy. Perovskite solar cell with a mix of CNT and CuSCN electrode exhibits the lowest series resistance of 76.69 Ω, resulting in the optimum solar cell performance such as a short-circuit current density (JSC) of 0.199 mA/cm2, open-circuit voltage (VOC) of 0.52 V, fill-factor (FF) of 0.369, and efficiency of 0.0735.
{"title":"Material Characteristics and Electrical Performance of Perovskite Solar Cells with Different Carbon-Based Electrodes Mixed with CuSCN","authors":"Elang Aji Defrianto, Atya Saniah, S. F. Rahman, N. R. Poespawati","doi":"10.1155/2023/8931693","DOIUrl":"https://doi.org/10.1155/2023/8931693","url":null,"abstract":"Perovskite solar cells are the most cutting-edge photovoltaic technology having high efficiency and short fabrication time. In recent decades, there has been a significant rise in the study of the usage of carbon materials in perovskite solar cells because of low cost and earth abundance. Several studies have been conducted to mix hole transport materials with carbon materials to improve the hole extraction capability. Nevertheless, no research has reported using CuSCN on different carbon electrodes on perovskite solar cells. In this research, various carbon materials, including carbon nanotubes (CNT), graphite, activated carbon, and reduced graphene oxide (rGO), are mixed with CuSCN. The carbon materials and CuSCN were mixed by ball mill and then deposited using the doctor blading method to become an electrode layer. The existence of CuSCN in carbon materials was proved by conducting the energy dispersive X-ray test. CNT mixed with CuSCN material exhibits the highest electrical conductivity indicated by ID/IG ratio of 1.22 using Raman spectroscopy. Perovskite solar cell with a mix of CNT and CuSCN electrode exhibits the lowest series resistance of 76.69 Ω, resulting in the optimum solar cell performance such as a short-circuit current density (JSC) of 0.199 mA/cm2, open-circuit voltage (VOC) of 0.52 V, fill-factor (FF) of 0.369, and efficiency of 0.0735.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77784490","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To address the problems of 5G network planning and optimization, a 5G user time delay prediction model based on the BiLSTM neural network optimized by APSO-SD is proposed. First, a channel generative model based on the ray-tracing model and the statistical channel model is constructed to obtain a large amount of time delay data, and a 5G user ray data feature model based on three-dimensional stereo mapping is proposed for input feature extraction. Then, an adaptive particle swarm optimization algorithm based on a search perturbation mechanism and differential enhancement strategy (APSO-SD) is proposed for the parameters’ optimization of BiLSTM neural networks. Finally, the APSO-SD-BiLSTM model is proposed to predict the time delay of 5G users. The experimental results show that the APSO-SD has a better convergence performance and optimization performance in benchmark function optimization compared with the other PSO algorithms, and the APSO-SD-BiLSTM model has better user time delay prediction accuracy in different scenarios.
{"title":"A Time Delay Prediction Model of 5G Users Based on the BiLSTM Neural Network Optimized by APSO-SD","authors":"Xiaozheng Dang, Di He, Cong Xie","doi":"10.1155/2023/4137614","DOIUrl":"https://doi.org/10.1155/2023/4137614","url":null,"abstract":"To address the problems of 5G network planning and optimization, a 5G user time delay prediction model based on the BiLSTM neural network optimized by APSO-SD is proposed. First, a channel generative model based on the ray-tracing model and the statistical channel model is constructed to obtain a large amount of time delay data, and a 5G user ray data feature model based on three-dimensional stereo mapping is proposed for input feature extraction. Then, an adaptive particle swarm optimization algorithm based on a search perturbation mechanism and differential enhancement strategy (APSO-SD) is proposed for the parameters’ optimization of BiLSTM neural networks. Finally, the APSO-SD-BiLSTM model is proposed to predict the time delay of 5G users. The experimental results show that the APSO-SD has a better convergence performance and optimization performance in benchmark function optimization compared with the other PSO algorithms, and the APSO-SD-BiLSTM model has better user time delay prediction accuracy in different scenarios.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84516910","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R. Chang, Bingzhen Zhang, Qianxin Zhu, Shan Zhao, Kai Yan, Y. Yang
Ensuring compliance with safety regulations regarding wearing is essential for the safety and security of those working on substation construction sites. However, relying on supervisors to monitor workers in real time on the work site or through remote surveillance videos is both unreasonable and inefficient. A deep learning network approach named FFA-YOLOv7 is presented in this study that utilizes an improved version of YOLOv7 to detect violations of worker wearing in real time during power construction site surveillance. In YOLOv7, the feature pyramid network (FPN) of the neck stage is constructed through continuous upsampling and skip connections for feature fusion, after continuous downsampling of the backbone. However, this process can result in the loss of precise shallow position information. To tackle this issue, we have introduced a novel feature fusion pathway to the FPN architecture, enabling each layer not only to fuse feature maps from the same level during the downsampling course but also to fuse feature maps from shallower levels. This approach combines precise positional information from shallow layers with rich semantic information from deep layers. Additionally, we utilized attention after feature fusion in each layer to optimize the feature map fusion effect and achieve better detection accuracy performance. In order to conduct comparative experiments, we trained six variations of the YOLO model as detectors using a dataset gathered from realistic construction sites. The experimental results indicate that our proposed FFA-YOLOv7 attained a detection precision of 95.92% and a recall rate of 97.13%, demonstrating a high level of accuracy and a low rate of missed detections. These outcomes effectively satisfy the requirements for robust and accurate detection of real-world power construction violations.
{"title":"FFA-YOLOv7: Improved YOLOv7 Based on Feature Fusion and Attention Mechanism for Wearing Violation Detection in Substation Construction Safety","authors":"R. Chang, Bingzhen Zhang, Qianxin Zhu, Shan Zhao, Kai Yan, Y. Yang","doi":"10.1155/2023/9772652","DOIUrl":"https://doi.org/10.1155/2023/9772652","url":null,"abstract":"Ensuring compliance with safety regulations regarding wearing is essential for the safety and security of those working on substation construction sites. However, relying on supervisors to monitor workers in real time on the work site or through remote surveillance videos is both unreasonable and inefficient. A deep learning network approach named FFA-YOLOv7 is presented in this study that utilizes an improved version of YOLOv7 to detect violations of worker wearing in real time during power construction site surveillance. In YOLOv7, the feature pyramid network (FPN) of the neck stage is constructed through continuous upsampling and skip connections for feature fusion, after continuous downsampling of the backbone. However, this process can result in the loss of precise shallow position information. To tackle this issue, we have introduced a novel feature fusion pathway to the FPN architecture, enabling each layer not only to fuse feature maps from the same level during the downsampling course but also to fuse feature maps from shallower levels. This approach combines precise positional information from shallow layers with rich semantic information from deep layers. Additionally, we utilized attention after feature fusion in each layer to optimize the feature map fusion effect and achieve better detection accuracy performance. In order to conduct comparative experiments, we trained six variations of the YOLO model as detectors using a dataset gathered from realistic construction sites. The experimental results indicate that our proposed FFA-YOLOv7 attained a detection precision of 95.92% and a recall rate of 97.13%, demonstrating a high level of accuracy and a low rate of missed detections. These outcomes effectively satisfy the requirements for robust and accurate detection of real-world power construction violations.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75221684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Normally, the buck-boost converter adopts single or double closed-loop control, and there are differences in control and parameters for different working modes and loads. In this study, a unified control method, the passivity-based control (PBC), is applied to a buck-boost converter for different loads, including constant resistance load (CRL), constant power load (CPL), and battery load. The PBC is a nonlinear control based on energy dissipation principle, and it has strong robustness to parameter interference and external disturbance, and it also has the advantages of simple design and simple implementation. Although many research studies have been conducted before, the voltage and current-related power losses are considered, and different load models are also compared in this research. The detailed mathematical model, control principle, and controller design of the buck-boost converter are thoroughly analysed. In addition, SIMULINK-based simulation results and experimental verification results of different loads are also given in the paper. Also, the PBC has smaller current overshot and smaller current ripples compared with PI control in different loads condition.
{"title":"Passivity-Based Control of Buck-Boost Converter for Different Loads Research","authors":"Feng Zhang, Jianguo Li, Gejun Zhu, Rongyuan Hu, Yaping Qu, Yujiang Zhang","doi":"10.1155/2023/5558246","DOIUrl":"https://doi.org/10.1155/2023/5558246","url":null,"abstract":"Normally, the buck-boost converter adopts single or double closed-loop control, and there are differences in control and parameters for different working modes and loads. In this study, a unified control method, the passivity-based control (PBC), is applied to a buck-boost converter for different loads, including constant resistance load (CRL), constant power load (CPL), and battery load. The PBC is a nonlinear control based on energy dissipation principle, and it has strong robustness to parameter interference and external disturbance, and it also has the advantages of simple design and simple implementation. Although many research studies have been conducted before, the voltage and current-related power losses are considered, and different load models are also compared in this research. The detailed mathematical model, control principle, and controller design of the buck-boost converter are thoroughly analysed. In addition, SIMULINK-based simulation results and experimental verification results of different loads are also given in the paper. Also, the PBC has smaller current overshot and smaller current ripples compared with PI control in different loads condition.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84038959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. V. G. Rao, M. Hema, Ramakrishna Raghutu, Ramakrishna S. S. Nuvvula, Polamarasetty P. Kumar, I. Colak, B. Khan
Stationary random-access memory (SRAM) undergoes an expansion stage, to repel advanced process variation and support ultra-low power operation. Memories occupy more than 80% of the surface in today’s microdevices, and this trend is expected to continue. Metal oxide semiconductor field effect transistor (MOSFET) face a set of difficulties, that results in higher leakage current (Ileakage) at lower strategy collisions. Fin field effect transistor (FinFET) is a highly effective substitute to complementary metal oxide semiconductor (CMOS) under the 45 nm variant due to advanced stability. Memory cells are significant in the large-scale computation system. SRAM is the most commonly used memory type; SRAMs are thought to utilize more than 60% of the chip area. The proposed SRAM cell is developed with FinFETs at 16 nm knot. Power, delay, power delay product (PDP), Ileakage, and stationary noise margin (SNM) are compared with traditional 6T SRAM cells. The designed cell decreases leakage power, current, and read access time. While comparing 6T SRAM and earlier low power SRAM cells, FinFET-based 10T SRAM provides significant SNM with reduced access time. The proposed 10T SRAM based on FinFET provides an 80.80% PDP reduction in write mode and a 50.65% PDP reduction in read mode compared to MOSEFET models. There is an improvement of 22.20% in terms of SNM and 25.53% in terms of Ileakage.
{"title":"Design and Development of Efficient SRAM Cell Based on FinFET for Low Power Memory Applications","authors":"M. V. G. Rao, M. Hema, Ramakrishna Raghutu, Ramakrishna S. S. Nuvvula, Polamarasetty P. Kumar, I. Colak, B. Khan","doi":"10.1155/2023/7069746","DOIUrl":"https://doi.org/10.1155/2023/7069746","url":null,"abstract":"Stationary random-access memory (SRAM) undergoes an expansion stage, to repel advanced process variation and support ultra-low power operation. Memories occupy more than 80% of the surface in today’s microdevices, and this trend is expected to continue. Metal oxide semiconductor field effect transistor (MOSFET) face a set of difficulties, that results in higher leakage current (Ileakage) at lower strategy collisions. Fin field effect transistor (FinFET) is a highly effective substitute to complementary metal oxide semiconductor (CMOS) under the 45 nm variant due to advanced stability. Memory cells are significant in the large-scale computation system. SRAM is the most commonly used memory type; SRAMs are thought to utilize more than 60% of the chip area. The proposed SRAM cell is developed with FinFETs at 16 nm knot. Power, delay, power delay product (PDP), Ileakage, and stationary noise margin (SNM) are compared with traditional 6T SRAM cells. The designed cell decreases leakage power, current, and read access time. While comparing 6T SRAM and earlier low power SRAM cells, FinFET-based 10T SRAM provides significant SNM with reduced access time. The proposed 10T SRAM based on FinFET provides an 80.80% PDP reduction in write mode and a 50.65% PDP reduction in read mode compared to MOSEFET models. There is an improvement of 22.20% in terms of SNM and 25.53% in terms of Ileakage.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-06-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81710869","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ali Othman Albaji, R. Rashid, Siti Zeleha Abdul Hamid
This project aims to investigate the best machine learning (ML) algorithm for classifying sounds originating from the environment that were considered noise pollution in smart cities. Sound collection was carried out using necessary sound capture tools, after which ML classification models were utilized for sound recognition. Additionally, noise pollution monitoring using Python was conducted to provide accurate results for sixteen different types of noise that were collected in sixteen cities in Malaysia. The numbers on the diagonal represent the correctly classified noises from the test set. Using these correlation matrices, the F1 score was calculated, and a comparison was performed for all models. The best model was found to be random forest.
{"title":"Investigation on Machine Learning Approaches for Environmental Noise Classifications","authors":"Ali Othman Albaji, R. Rashid, Siti Zeleha Abdul Hamid","doi":"10.1155/2023/3615137","DOIUrl":"https://doi.org/10.1155/2023/3615137","url":null,"abstract":"This project aims to investigate the best machine learning (ML) algorithm for classifying sounds originating from the environment that were considered noise pollution in smart cities. Sound collection was carried out using necessary sound capture tools, after which ML classification models were utilized for sound recognition. Additionally, noise pollution monitoring using Python was conducted to provide accurate results for sixteen different types of noise that were collected in sixteen cities in Malaysia. The numbers on the diagonal represent the correctly classified noises from the test set. Using these correlation matrices, the F1 score was calculated, and a comparison was performed for all models. The best model was found to be random forest.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76894774","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The ionospheric state is becoming increasingly important to forecast for the reliable operation of terrestrial and space-based radio-communication systems which are influenced by ionospheric space weather. In this study, we have investigated and tested a multivariate long short-term memory (LSTM) deep learning model for its forecasting accuracy over different latitudinal regions during the solar quiet and solar active years. We also tested its prediction capability during the occurrence of a geomagnetic storm. Four stations qaq1 (60.7°N, 46.04°W), baie (49.18°N, 68.26°W), mas1 (27.76°N, 15.63°W), and bogt (4.64°N, 74.08°W) in the northern hemisphere were used in this study. To optimize the feature extraction process, we used heat map to find the correlation between TEC and the various exogenous parameters and finally nine correlated parameters were used as inputs to train the LSTM model. The performance of the LSTM model was validated by comparing it with the multilayer perceptron (MLP) machine learning algorithm using root mean square error (RMSE) and mean absolute error (MAE) as evaluation indices. The results showed an accuracy improvement of 70% and 64% over MLP during the solar quiet and active years, respectively. The prediction accuracy of our LSTM model was also 74% better than MLP during the geomagnetic storm event. These findings demonstrate the effectiveness of the developed LSTM model and the right selection of the exogenous parameters in estimating TEC, and suggest that this LSTM model can be used for short-term TEC forecasting.
{"title":"Forecasting of Ionospheric Total Electron Content Data Using Multivariate Deep LSTM Model for Different Latitudes and Solar Activity","authors":"Nayana Shenvi, Hassanali Virani","doi":"10.1155/2023/2855762","DOIUrl":"https://doi.org/10.1155/2023/2855762","url":null,"abstract":"The ionospheric state is becoming increasingly important to forecast for the reliable operation of terrestrial and space-based radio-communication systems which are influenced by ionospheric space weather. In this study, we have investigated and tested a multivariate long short-term memory (LSTM) deep learning model for its forecasting accuracy over different latitudinal regions during the solar quiet and solar active years. We also tested its prediction capability during the occurrence of a geomagnetic storm. Four stations qaq1 (60.7°N, 46.04°W), baie (49.18°N, 68.26°W), mas1 (27.76°N, 15.63°W), and bogt (4.64°N, 74.08°W) in the northern hemisphere were used in this study. To optimize the feature extraction process, we used heat map to find the correlation between TEC and the various exogenous parameters and finally nine correlated parameters were used as inputs to train the LSTM model. The performance of the LSTM model was validated by comparing it with the multilayer perceptron (MLP) machine learning algorithm using root mean square error (RMSE) and mean absolute error (MAE) as evaluation indices. The results showed an accuracy improvement of 70% and 64% over MLP during the solar quiet and active years, respectively. The prediction accuracy of our LSTM model was also 74% better than MLP during the geomagnetic storm event. These findings demonstrate the effectiveness of the developed LSTM model and the right selection of the exogenous parameters in estimating TEC, and suggest that this LSTM model can be used for short-term TEC forecasting.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90511291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tingzhong Wang, Yongxin Zhang, Yifan Zhang, Hao Lu, Bo Yu, Shoubo Peng, Youzhong Ma, Deguang Li
The typical pretrained model’s feature extraction capabilities are insufficient for medical named entity identification, and it is challenging to express word polysemy, resulting in a low recognition accuracy for electronic medical records. In order to solve this problem, this paper proposes a new model that combines the BERT pretraining model and the BilSTM-CRF model. First, word embedding with semantic information is obtained by pretraining the corpus input to the BERT model. Then, the BiLSTM module is utilized to extract further features from the encoded outputs of BERT in order to account for context information and improve the accuracy of semantic coding. Then, CRF is used to modify the results of BiLSTM to screen out the annotation sequence with the largest score. Finally, extensive experimental results show that the performance of the proposed model is effectively improved compared with other models.
{"title":"A Hybrid Model Based on Deep Convolutional Network for Medical Named Entity Recognition","authors":"Tingzhong Wang, Yongxin Zhang, Yifan Zhang, Hao Lu, Bo Yu, Shoubo Peng, Youzhong Ma, Deguang Li","doi":"10.1155/2023/8969144","DOIUrl":"https://doi.org/10.1155/2023/8969144","url":null,"abstract":"The typical pretrained model’s feature extraction capabilities are insufficient for medical named entity identification, and it is challenging to express word polysemy, resulting in a low recognition accuracy for electronic medical records. In order to solve this problem, this paper proposes a new model that combines the BERT pretraining model and the BilSTM-CRF model. First, word embedding with semantic information is obtained by pretraining the corpus input to the BERT model. Then, the BiLSTM module is utilized to extract further features from the encoded outputs of BERT in order to account for context information and improve the accuracy of semantic coding. Then, CRF is used to modify the results of BiLSTM to screen out the annotation sequence with the largest score. Finally, extensive experimental results show that the performance of the proposed model is effectively improved compared with other models.","PeriodicalId":23352,"journal":{"name":"Turkish J. Electr. Eng. Comput. Sci.","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-05-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73281582","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：