In recent years with the advent of computational power, Machine Learning has become a popular approach in financial forecasting, particularly for stock price analysis. In this paper, the authors develop a non-recurrent active trading algorithm based on stock price prediction, using Support Vector Machines on high frequency data, and compare its risk adjusted performance to the returns of a statistical portfolio predicted by the Capital Asset Pricing Model. The authors selected the three highest volume securities from a pool of 100 initially selected stock dataset to investigate the algorithmic trading strategy. The abnormal return estimates are significant and positive, and the systematic risk is lower than unity in all cases, suggesting lower risk compared to the market. Moreover, the estimated beta values for all stocks were close to zero, indicating a market independent process. The correlation analysis revealed weak correlations among the processes, supporting the potential for risk reduction and volatility mitigation through portfolio diversification. The authors tested an equally weighted portfolio of the selected three assets and demonstrated a remarkable return of 1348% during the evaluation period from July 1st, 2020, to January 1st, 2023. The results suggest that the weak form of market efficiency can be questioned, as the algorithmic trading strategy, employing a Support Vector Machine binary classification model, has consistently generated statistically significant and substantial abnormal returns using historical market data.
{"title":"EVALUATION OF SUPPORT VECTOR MACHINE BASED STOCK PRICE PREDICTION","authors":"Tilla IZSÁK, László MARÁK, Mihály ORMOS","doi":"10.35784/acs-2023-25","DOIUrl":"https://doi.org/10.35784/acs-2023-25","url":null,"abstract":"
 
 
 In recent years with the advent of computational power, Machine Learning has become a popular approach in financial forecasting, particularly for stock price analysis. In this paper, the authors develop a non-recurrent active trading algorithm based on stock price prediction, using Support Vector Machines on high frequency data, and compare its risk adjusted performance to the returns of a statistical portfolio predicted by the Capital Asset Pricing Model. The authors selected the three highest volume securities from a pool of 100 initially selected stock dataset to investigate the algorithmic trading strategy. The abnormal return estimates are significant and positive, and the systematic risk is lower than unity in all cases, suggesting lower risk compared to the market. Moreover, the estimated beta values for all stocks were close to zero, indicating a market independent process. The correlation analysis revealed weak correlations among the processes, supporting the potential for risk reduction and volatility mitigation through portfolio diversification. The authors tested an equally weighted portfolio of the selected three assets and demonstrated a remarkable return of 1348% during the evaluation period from July 1st, 2020, to January 1st, 2023. The results suggest that the weak form of market efficiency can be questioned, as the algorithmic trading strategy, employing a Support Vector Machine binary classification model, has consistently generated statistically significant and substantial abnormal returns using historical market data.
 
 
","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135126784","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}
Sri INDRA MAIYANTI, Anita DESIANI, Syafrina LAMIN, P PUSPITAHATI, Muhammad ARHAMI, Nuni GOFAR, Destika CAHYANA
Soil is a solid-particle that covers the earth's surface. Soils can be classified based their color. The color can be an indication of soil properties and soil conditions. Soil image classification requires high accuracy and caution. CNN works well on image classification, but CNN requires a large amount of data. Augmentation is one technique to overcome data needs like rotation and improving contrast. Rotation is the movement of rotating the image position randomly to various degrees. Gamma Correction is a method to improve image by decreasing or increasing the contrast. The rotation and Gamma Correction on augmentation can increase the amount of training data from 156 to 2500 soil images data. The classification of soil data is not referred to soil taxonomy system such as Entisols and Histosols but it used arbitrary simple classification based on color. Unfortunately, the weakness of the CNN is vanishing and exploded gradients. Another Deep learning that can overcome vanishing and exploded gradients is dense blocks. This study proposes a combination of Augmentation and CNN-Dense block where in the augmentation a combination of rotation and Gamma-correction techniques is used and Soil image classification based on color is used by the CNN-Dense block. The combination method is able to give excellent results, where all performances accuracy, precisions, recall and F1-Score are above 90%. The combination of rotation and Gamma Correction on augmentation and CNN is a robust method to use in soil image classification based on color.
{"title":"ROTATION-GAMMA CORRECTION AUGMENTATION ON CNN-DENSE BLOCK FOR SOIL IMAGE CLASSIFICATION","authors":"Sri INDRA MAIYANTI, Anita DESIANI, Syafrina LAMIN, P PUSPITAHATI, Muhammad ARHAMI, Nuni GOFAR, Destika CAHYANA","doi":"10.35784/acs-2023-27","DOIUrl":"https://doi.org/10.35784/acs-2023-27","url":null,"abstract":"Soil is a solid-particle that covers the earth's surface. Soils can be classified based their color. The color can be an indication of soil properties and soil conditions. Soil image classification requires high accuracy and caution. CNN works well on image classification, but CNN requires a large amount of data. Augmentation is one technique to overcome data needs like rotation and improving contrast. Rotation is the movement of rotating the image position randomly to various degrees. Gamma Correction is a method to improve image by decreasing or increasing the contrast. The rotation and Gamma Correction on augmentation can increase the amount of training data from 156 to 2500 soil images data. The classification of soil data is not referred to soil taxonomy system such as Entisols and Histosols but it used arbitrary simple classification based on color. Unfortunately, the weakness of the CNN is vanishing and exploded gradients. Another Deep learning that can overcome vanishing and exploded gradients is dense blocks. This study proposes a combination of Augmentation and CNN-Dense block where in the augmentation a combination of rotation and Gamma-correction techniques is used and Soil image classification based on color is used by the CNN-Dense block. The combination method is able to give excellent results, where all performances accuracy, precisions, recall and F1-Score are above 90%. The combination of rotation and Gamma Correction on augmentation and CNN is a robust method to use in soil image classification based on color.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135127142","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}
Face recognition technology has been widely used in all aspects of people's lives. However, the accuracy of face recognition is greatly reduced due to the obscuring of objects, such as masks and sunglasses. Wearing masks in public has been a crucial approach to preventing illness, especially since the Covid-19 outbreak. This poses challenges to applications such as face recognition. Therefore, the removal of masks via image inpainting has become a hot topic in the field of computer vision. Deep learning-based image inpainting techniques have taken observable results, but the restored images still have problems such as blurring and inconsistency. To address such problems, this paper proposes an improved inpainting model based on generative adversarial network: the model adds attention mechanisms to the sampling module based on pix2pix network; the residual module is improved by adding convolutional branches. The improved inpainting model can not only effectively restore faces obscured by face masks, but also realize the inpainting of randomly obscured images of human faces. To further validate the generality of the inpainting model, tests are conducted on the datasets of CelebA, Paris Street and Place2, and the experimental results show that both SSIM and PSNR have improved significantly.
{"title":"MASK FACE INPAINTING BASED ON IMPROVED GENERATIVE ADVERSARIAL NETWORK","authors":"Qingyun Liu, Roben A. Juanatas","doi":"10.35784/acs-2023-12","DOIUrl":"https://doi.org/10.35784/acs-2023-12","url":null,"abstract":"Face recognition technology has been widely used in all aspects of people's lives. However, the accuracy of face recognition is greatly reduced due to the obscuring of objects, such as masks and sunglasses. Wearing masks in public has been a crucial approach to preventing illness, especially since the Covid-19 outbreak. This poses challenges to applications such as face recognition. Therefore, the removal of masks via image inpainting has become a hot topic in the field of computer vision. Deep learning-based image inpainting techniques have taken observable results, but the restored images still have problems such as blurring and inconsistency. To address such problems, this paper proposes an improved inpainting model based on generative adversarial network: the model adds attention mechanisms to the sampling module based on pix2pix network; the residual module is improved by adding convolutional branches. The improved inpainting model can not only effectively restore faces obscured by face masks, but also realize the inpainting of randomly obscured images of human faces. To further validate the generality of the inpainting model, tests are conducted on the datasets of CelebA, Paris Street and Place2, and the experimental results show that both SSIM and PSNR have improved significantly.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48773308","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}
In our aging culture, neurodegenerative disorders like Parkinson's disease (PD) are among the most serious health issues. It is a neurological condition that has social and economic effects on individuals. It happens because the brain's dopamine-producing cells are unable to produce enough of the chemical to support the body's motor functions. The main symptoms of this illness are eyesight, excretion activity, speech, and mobility issues, followed by depression, anxiety, sleep issues, and panic attacks. The main aim of this research is to develop a workable clinical decision-making framework that aids the physician in diagnosing patients with PD influence. In this research, we proposed a technique to classify Parkinson’s disease by MRI brain images. Initially, normalize the input data using the min-max normalization method and then remove noise from input images using a median filter. Then utilizing the Binary Dragonfly Algorithm to select the features. Furthermore, to segment the diseased part from MRI brain images using the technique Dense-UNet. Then, classify the disease as if it’s Parkinson’s disease or health control using the Deep Residual Convolutional Neural Network (DRCNN) technique along with Enhanced Whale Optimization Algorithm (EWOA) to get better classification accuracy. Here, we use the public Parkinson’s Progression Marker Initiative (PPMI) dataset for Parkinson’s MRI images. The accuracy, sensitivity, specificity, and precision metrics will be utilized with manually gathered data to assess the efficacy of the proposed methodology.
{"title":"CLASSIFICATION OF PARKINSON'S DISEASE IN BRAIN MRI IMAGES USING DEEP RESIDUAL CONVOLUTIONAL NEURAL NETWORK","authors":"Puppala Praneeth, Majety Sathvika, Vivek Kommareddy, Madala Sarath, Saran Mallela, Koneru Suvarna Vani, Prasun Chkrabarti","doi":"10.35784/acs-2023-19","DOIUrl":"https://doi.org/10.35784/acs-2023-19","url":null,"abstract":"In our aging culture, neurodegenerative disorders like Parkinson's disease (PD) are among the most serious health issues. It is a neurological condition that has social and economic effects on individuals. It happens because the brain's dopamine-producing cells are unable to produce enough of the chemical to support the body's motor functions. The main symptoms of this illness are eyesight, excretion activity, speech, and mobility issues, followed by depression, anxiety, sleep issues, and panic attacks. The main aim of this research is to develop a workable clinical decision-making framework that aids the physician in diagnosing patients with PD influence. In this research, we proposed a technique to classify Parkinson’s disease by MRI brain images. Initially, normalize the input data using the min-max normalization method and then remove noise from input images using a median filter. Then utilizing the Binary Dragonfly Algorithm to select the features. Furthermore, to segment the diseased part from MRI brain images using the technique Dense-UNet. Then, classify the disease as if it’s Parkinson’s disease or health control using the Deep Residual Convolutional Neural Network (DRCNN) technique along with Enhanced Whale Optimization Algorithm (EWOA) to get better classification accuracy. Here, we use the public Parkinson’s Progression Marker Initiative (PPMI) dataset for Parkinson’s MRI images. The accuracy, sensitivity, specificity, and precision metrics will be utilized with manually gathered data to assess the efficacy of the proposed methodology.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43658001","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}
Intelligent mobile robots must possess the ability to navigate in complex environments. The field of mobile robot navigation is continuously evolving, with various technologies being developed. Deep learning has gained attention from researchers, and numerous navigation models utilizing deep learning have been proposed. In this study, the YOLOv5 model is utilized to identify objects to aid the mobile robot in determining movement conditions. However, the limitation of deep learning models being trained on insufficient data, leading to inaccurate recognition in unforeseen scenarios, is addressed by introducing an innovative computer vision technology that detects lanes in real-time. Combining the deep learning model with computer vision technology, the robot can identify different types of objects, allowing it to estimate distance and adjust speed accordingly. Additionally, the paper investigates the recognition reliability in varying light intensities. The findings of this study offer promising directions for future breakthroughs in mobile robot navigation
{"title":"NAVIGATION STRATEGY FOR MOBILE ROBOT BASED ON COMPUTER VISION AND YOLOV5 NETWORK IN THE UNKNOWN ENVIRONMENT","authors":"Thanh-Lam Bui, Ngoc-Tien Tran","doi":"10.35784/acs-2023-16","DOIUrl":"https://doi.org/10.35784/acs-2023-16","url":null,"abstract":"Intelligent mobile robots must possess the ability to navigate in complex environments. The field of mobile robot navigation is continuously evolving, with various technologies being developed. Deep learning has gained attention from researchers, and numerous navigation models utilizing deep learning have been proposed. In this study, the YOLOv5 model is utilized to identify objects to aid the mobile robot in determining movement conditions. However, the limitation of deep learning models being trained on insufficient data, leading to inaccurate recognition in unforeseen scenarios, is addressed by introducing an innovative computer vision technology that detects lanes in real-time. Combining the deep learning model with computer vision technology, the robot can identify different types of objects, allowing it to estimate distance and adjust speed accordingly. Additionally, the paper investigates the recognition reliability in varying light intensities. The findings of this study offer promising directions for future breakthroughs in mobile robot navigation","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48586832","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 main aim of predictive maintenance is to minimize downtime, failure risks and maintenance costs in manufacturing systems. Over the past few years, machine learning methods gained ground with diverse and successful applications in the area of predictive maintenance. This study shows that performing preprocessing techniques such as oversampling and features selection for failure prediction, is promising. For instance, to handle imbalanced data, the SMOTE-Tomek method is used. For features selection, three different methods can be applied: Recursive Feature Elimination, Random Forest and Variance Threshold. The data considered in this paper for simulation is used in literature; it is applied to aircraft engine sensors measurements to predict engines failure, while the predicting algorithm used is a Support Vector Machine. The results show that classification accuracy can be significantly boosted by using the preprocessing techniques.
{"title":"HYBRID FEATURE SELECTION AND SUPPORT VECTOR MACHINE FRAMEWORK FOR PREDICTING MAINTENANCE FAILURES","authors":"Mouna Tarik, Ayoub Mniai, K. Jebari","doi":"10.35784/acs-2023-18","DOIUrl":"https://doi.org/10.35784/acs-2023-18","url":null,"abstract":"The main aim of predictive maintenance is to minimize downtime, failure risks and maintenance costs in manufacturing systems. Over the past few years, machine learning methods gained ground with diverse and successful applications in the area of predictive maintenance. This study shows that performing preprocessing techniques such as oversampling and features selection for failure prediction, is promising. For instance, to handle imbalanced data, the SMOTE-Tomek method is used. For features selection, three different methods can be applied: Recursive Feature Elimination, Random Forest and Variance Threshold. The data considered in this paper for simulation is used in literature; it is applied to aircraft engine sensors measurements to predict engines failure, while the predicting algorithm used is a Support Vector Machine. The results show that classification accuracy can be significantly boosted by using the preprocessing techniques.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45807515","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. Larios, Perfecto M. QUINTERO-FLORES, M. Anzures-García, Miguel CAMACHO-HERNANDEZ
This paper presents the application of a task scheduling algorithm called Fan on an artificial intelligence technique as genetic algorithms for the problem of finding minima in objective functions, where the equations are predefined to measure the return on an investment. This work combines the methodologies of exploration and exploitation of a population, obtaining results with good aptitudes until finding a better learning based on conditions of not ending until an individual delivers a better aptitude, complying with the established restrictions, exhausting all possible options and fulfilling a stop condition. A real-time task planning algorithm was applied based on consensus techniques. A software tool was developed, and the scheduler called FAN was adapted that contemplates the execution of periodic, aperiodic, and sporadic tasks focused on controlled environments, considering that strict time restrictions are met. In the first phase of the work, it is shown how convergence precipitates to an evolution, this is done in few iterations. In a second stage, exploitation was improved, giving the algorithm a better performance in convergence and feasibility. As a result, there is the exploitation of the population and applying iterations with the fan algorithm and better aptitudes were obtained that occur through asynchronized processes under real-time planning concurrently.
{"title":"APPLICATION OF THE REAL-TIME FAN SCHEDULING IN THE EXPLORATION-EXPLOITATION TO OPTIMIZE MINIMUM FUNCTIONS OBJECTIVES","authors":"M. Larios, Perfecto M. QUINTERO-FLORES, M. Anzures-García, Miguel CAMACHO-HERNANDEZ","doi":"10.35784/acs-2023-13","DOIUrl":"https://doi.org/10.35784/acs-2023-13","url":null,"abstract":"This paper presents the application of a task scheduling algorithm called Fan on an artificial intelligence technique as genetic algorithms for the problem of finding minima in objective functions, where the equations are predefined to measure the return on an investment. This work combines the methodologies of exploration and exploitation of a population, obtaining results with good aptitudes until finding a better learning based on conditions of not ending until an individual delivers a better aptitude, complying with the established restrictions, exhausting all possible options and fulfilling a stop condition. A real-time task planning algorithm was applied based on consensus techniques. A software tool was developed, and the scheduler called FAN was adapted that contemplates the execution of periodic, aperiodic, and sporadic tasks focused on controlled environments, considering that strict time restrictions are met. In the first phase of the work, it is shown how convergence precipitates to an evolution, this is done in few iterations. In a second stage, exploitation was improved, giving the algorithm a better performance in convergence and feasibility. As a result, there is the exploitation of the population and applying iterations with the fan algorithm and better aptitudes were obtained that occur through asynchronized processes under real-time planning concurrently.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41776116","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}
This paper presents a new method for generating nonlinear helical spring geometries based on a rigorous mathematical formulation. The model was developed for two scenarios for modifying a spring with a stepped helix angle: for a fixed helix angle of the active coils and for a fixed overall height of the spring. It allows the development of compression spring geometries with non-linear load-deflection curves, while maintaining predetermined values of selected geometrical parameters such as the number of passive and active coils and the total height or helix angle of the linear segment of the active coils. Based on the proposed models, Python scripts were developed that can be implemented in any CAD software offering scripting capabilities or equipped with Application Programming Interfaces. Examples of scripts that use the developed model to generate the geometry of selected springs are presented. FEM analyses of quasi-static compression tests carried out for these spring models have shown that, using the proposed tools, springs with a wide range of variation in static load-deflection curves can be obtained, including progressive springs with a high degree of nonlinearity in the characteristics. The obtained load-deflection curves can be described with a high degree of accuracy by power function. The proposed method can find applications in both machine design and spring manufacturing.
{"title":"A NEW METHOD FOR GENERATING VIRTUAL MODELS OF NONLINEAR HELICAL SPRINGS BASED ON A RIGOROUS MATHEMATICAL MODEL","authors":"K. Michalczyk, M. Warzecha, R. Baran","doi":"10.35784/acs-2023-17","DOIUrl":"https://doi.org/10.35784/acs-2023-17","url":null,"abstract":"This paper presents a new method for generating nonlinear helical spring geometries based on a rigorous mathematical formulation. The model was developed for two scenarios for modifying a spring with a stepped helix angle: for a fixed helix angle of the active coils and for a fixed overall height of the spring. It allows the development of compression spring geometries with non-linear load-deflection curves, while maintaining predetermined values of selected geometrical parameters such as the number of passive and active coils and the total height or helix angle of the linear segment of the active coils. Based on the proposed models, Python scripts were developed that can be implemented in any CAD software offering scripting capabilities or equipped with Application Programming Interfaces. Examples of scripts that use the developed model to generate the geometry of selected springs are presented. FEM analyses of quasi-static compression tests carried out for these spring models have shown that, using the proposed tools, springs with a wide range of variation in static load-deflection curves can be obtained, including progressive springs with a high degree of nonlinearity in the characteristics. The obtained load-deflection curves can be described with a high degree of accuracy by power function. The proposed method can find applications in both machine design and spring manufacturing.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44326028","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}
Parkinson's disease is a recognizable clinical syndrome with a variety of causes and clinical presentations; it represents a rapidly growing neurodegenerative disorder. Since about 90 percent of Parkinson's disease sufferers have some form of early speech impairment, recent studies on tele diagnosis of Parkinson's disease have focused on the recognition of voice impairments from vowel phonations or the subjects' discourse. In this paper, we present a new approach for Parkinson's disease detection from speech sounds that are based on CNN and LSTM and uses two categories of characteristics Mel Frequency Cepstral Coefficients (MFCC) and Gammatone Cepstral Coefficients (GTCC) obtained from noise-removed speech signals with comparative EMD-DWT and DWT-EMD analysis. The proposed model is divided into three stages. In the first step, noise is removed from the signals using the EMD-DWT and DWT-EMD methods. In the second step, the GTCC and MFCC are extracted from the enhanced audio signals. The classification process is carried out in the third step by feeding these features into the LSTM and CNN models, which are designed to define sequential information from the extracted features. The experiments are performed using PC-GITA and Sakar datasets and 10-fold cross validation method, the highest classification accuracy for the Sakar dataset reached 100% for both EMD-DWT-GTCC-CNN and DWT-EMD-GTCC-CNN, and for the PC-GITA dataset, the accuracy is reached 100% for EMD-DWT-GTCC-CNN and 96.55% for DWT-EMD-GTCC-CNN. The results of this study indicate that the characteristics of GTCC are more appropriate and accurate for the assessment of PD than MFCC.
{"title":"CNN AND LSTM FOR THE CLASSIFICATION OF PARKINSON'S DISEASE BASED ON THE GTCC AND MFCC","authors":"N. Boualoulou, T. BELHOUSSINE DRISSI, B. Nsiri","doi":"10.35784/acs-2023-11","DOIUrl":"https://doi.org/10.35784/acs-2023-11","url":null,"abstract":"Parkinson's disease is a recognizable clinical syndrome with a variety of causes and clinical presentations; it represents a rapidly growing neurodegenerative disorder. Since about 90 percent of Parkinson's disease sufferers have some form of early speech impairment, recent studies on tele diagnosis of Parkinson's disease have focused on the recognition of voice impairments from vowel phonations or the subjects' discourse. In this paper, we present a new approach for Parkinson's disease detection from speech sounds that are based on CNN and LSTM and uses two categories of characteristics Mel Frequency Cepstral Coefficients (MFCC) and Gammatone Cepstral Coefficients (GTCC) obtained from noise-removed speech signals with comparative EMD-DWT and DWT-EMD analysis. The proposed model is divided into three stages. In the first step, noise is removed from the signals using the EMD-DWT and DWT-EMD methods. In the second step, the GTCC and MFCC are extracted from the enhanced audio signals. The classification process is carried out in the third step by feeding these features into the LSTM and CNN models, which are designed to define sequential information from the extracted features. The experiments are performed using PC-GITA and Sakar datasets and 10-fold cross validation method, the highest classification accuracy for the Sakar dataset reached 100% for both EMD-DWT-GTCC-CNN and DWT-EMD-GTCC-CNN, and for the PC-GITA dataset, the accuracy is reached 100% for EMD-DWT-GTCC-CNN and 96.55% for DWT-EMD-GTCC-CNN. The results of this study indicate that the characteristics of GTCC are more appropriate and accurate for the assessment of PD than MFCC.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":" ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"47822139","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 purpose of this paper was to investigate in practice the possibility of using evolutionary algorithms to solve the traveling salesman problem on a real example. The goal was achieved by developing an original implementation of the evolutionary algorithm in Python, and by preparing an example of the traveling salesman problem in the form of a directed graph representing polish voivodship cities. As part of the work an application in Python was written. It provides a user interface which allows setting selected parameters of the evolutionary algorithm and solving the prepared problem. The results are presented in both text and graphical form. The correctness of the evolutionary algorithm's operation and the implementation was confirmed by performed tests. A large number of tested solutions (2500) and the analysis of the obtained results allowed for a conclusion that an optimal (relatively suboptimal) solution had been found.
{"title":"APPLICATION OF GENETIC ALGORITHMS TO THE TRAVELING SALESMAN PROBLEM","authors":"Tomasz D. Sikora, W. Gryglewicz-Kacerka","doi":"10.35784/acs-2023-14","DOIUrl":"https://doi.org/10.35784/acs-2023-14","url":null,"abstract":"The purpose of this paper was to investigate in practice the possibility of using evolutionary algorithms to solve the traveling salesman problem on a real example. The goal was achieved by developing an original implementation of the evolutionary algorithm in Python, and by preparing an example of the traveling salesman problem in the form of a directed graph representing polish voivodship cities. As part of the work an application in Python was written. It provides a user interface which allows setting selected parameters of the evolutionary algorithm and solving the prepared problem. The results are presented in both text and graphical form. The correctness of the evolutionary algorithm's operation and the implementation was confirmed by performed tests. A large number of tested solutions (2500) and the analysis of the obtained results allowed for a conclusion that an optimal (relatively suboptimal) solution had been found.","PeriodicalId":36379,"journal":{"name":"Applied Computer Science","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2023-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69999889","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：