Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936466
M. Mohan, Anuradha Patil, S. Mohana, P. Subhashini, Sumit Kushwaha, S. M. Pandian
Denoising magnetic resonance images are traditionally done individually, introducing undesired artefacts like blurring. To solve this issue, this paper offers a unique adaptive interpolation architecture that simultaneously allows for image data augmentation, noise removal, and detail augmentation. The multi-tier kernel (MTK) algorithm adjusts the extrapolation weights based on mathematical features in magnetic resonance (MR) data. The MTK weight matrix is then adaptively sharpened, and a Rician bias corrective is used to reduce Rician noise and improve small details. After processing, the noise eliminates the bias produced by the asymmetric sources. The adaptive MTK, in this way, extends the zero ordering estimating methodology to higher regression power facilitating edge maintenance and detail restoration. Investigation outcomes using genuine and MR images (with/without noise) evidenced that the algorithm delivered good restoration outcomes than conventional deep-learning-based approaches but with fewer requirements and calculation burden.
{"title":"Multi-Tier Kernel for Disease Prediction using Texture Analysis with MR Images","authors":"M. Mohan, Anuradha Patil, S. Mohana, P. Subhashini, Sumit Kushwaha, S. M. Pandian","doi":"10.1109/ICECAA55415.2022.9936466","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936466","url":null,"abstract":"Denoising magnetic resonance images are traditionally done individually, introducing undesired artefacts like blurring. To solve this issue, this paper offers a unique adaptive interpolation architecture that simultaneously allows for image data augmentation, noise removal, and detail augmentation. The multi-tier kernel (MTK) algorithm adjusts the extrapolation weights based on mathematical features in magnetic resonance (MR) data. The MTK weight matrix is then adaptively sharpened, and a Rician bias corrective is used to reduce Rician noise and improve small details. After processing, the noise eliminates the bias produced by the asymmetric sources. The adaptive MTK, in this way, extends the zero ordering estimating methodology to higher regression power facilitating edge maintenance and detail restoration. Investigation outcomes using genuine and MR images (with/without noise) evidenced that the algorithm delivered good restoration outcomes than conventional deep-learning-based approaches but with fewer requirements and calculation burden.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115371422","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936336
Chen Chee Kin, Zailan Arabee Bin Abdul Salam, Kadhar Batcha Nowshath
In this digital era, People have become more aware on purchasing a new property. Many digital tools have been developed to analyze the property marketing strategies and the buyers' budget constraints. The goal of this paper is to predict house prices for non-home owners based on their financial resources and aspirations. Estimated prices will be calculated by using different tools such as Machine Learning (ML), Artificial Neural Network (ANN) and Chatbot. All of the above-mentioned techniques were used here to determine the most effective house price from the collected dataset. This research project will particularly conduct multiple researches on the affordability of houses present within Malaysia. The motive of this work is to build a prediction model to help in the process of house price prediction and assist both buyers and seller to have a general view on the current market price and trend.
{"title":"Machine Learning based House Price Prediction Model","authors":"Chen Chee Kin, Zailan Arabee Bin Abdul Salam, Kadhar Batcha Nowshath","doi":"10.1109/ICECAA55415.2022.9936336","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936336","url":null,"abstract":"In this digital era, People have become more aware on purchasing a new property. Many digital tools have been developed to analyze the property marketing strategies and the buyers' budget constraints. The goal of this paper is to predict house prices for non-home owners based on their financial resources and aspirations. Estimated prices will be calculated by using different tools such as Machine Learning (ML), Artificial Neural Network (ANN) and Chatbot. All of the above-mentioned techniques were used here to determine the most effective house price from the collected dataset. This research project will particularly conduct multiple researches on the affordability of houses present within Malaysia. The motive of this work is to build a prediction model to help in the process of house price prediction and assist both buyers and seller to have a general view on the current market price and trend.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114372655","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936488
M. Krishna, J. Praveenchandar
The study aims to identify the frauds committed using a payment card such as credit cards, debit cards, and also an experiment is performed to find the best suitable algorithm among Random forest and Logistic Regression. Materials and Methods: To stop the fraud detections using Random forest (N=10) and Logistic regression (N=10) with supervised learning that gives insights from the previous data. Results: The precision of the random forest is 76.29% compared with Logistic regression with accuracy of 74.65% with statistical significance value p=0.03 (p<0.05) using Independent sample t test. Conclusion: This results proved that Random forest was significantly better for Fraud detection than Logistic regression within the study’s limits.
{"title":"Comparative Analysis of Credit Card Fraud Detection using Logistic regression with Random Forest towards an Increase in Accuracy of Prediction","authors":"M. Krishna, J. Praveenchandar","doi":"10.1109/ICECAA55415.2022.9936488","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936488","url":null,"abstract":"The study aims to identify the frauds committed using a payment card such as credit cards, debit cards, and also an experiment is performed to find the best suitable algorithm among Random forest and Logistic Regression. Materials and Methods: To stop the fraud detections using Random forest (N=10) and Logistic regression (N=10) with supervised learning that gives insights from the previous data. Results: The precision of the random forest is 76.29% compared with Logistic regression with accuracy of 74.65% with statistical significance value p=0.03 (p<0.05) using Independent sample t test. Conclusion: This results proved that Random forest was significantly better for Fraud detection than Logistic regression within the study’s limits.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"167 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114662402","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936202
A. Nithya, K. Ishwarya, Guneet Mummaneni, Vaibhavi Verma
Human Activity Recognition has gained greater emphasize in the last few years due to its widespread applicability and psychological curiosity. This system can be adopted in innumerable applications, like healthcare monitoring systems, surveillance systems, and so on. Smart-phones have built-in multifunctional sensors like as accelerometers and gyroscopes that provide useful sensory data when participants perform daily activities thus helping in HAR activity. Highly efficient features are extracted from this sensor data and techniques like denoising, normalization and segmentation are used to reduce noise and extract valuable feature vectors. Prior research showed that deep learning methods like recurrent neural networks and one-dimensional convolution networks provide excellent results in activity recognition tasks. In this paper, an ensemble model of CNN and SVM is proposed to further improve the accuracy and provide a robust model. Experimental methods are tested on UCI-HAR dataset and compared with other state-of-the-art methods like LSTM, CNN-LSTM, and Conv LSTM.
{"title":"CNN based Identifying Human Activity using Smartphone Sensors","authors":"A. Nithya, K. Ishwarya, Guneet Mummaneni, Vaibhavi Verma","doi":"10.1109/ICECAA55415.2022.9936202","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936202","url":null,"abstract":"Human Activity Recognition has gained greater emphasize in the last few years due to its widespread applicability and psychological curiosity. This system can be adopted in innumerable applications, like healthcare monitoring systems, surveillance systems, and so on. Smart-phones have built-in multifunctional sensors like as accelerometers and gyroscopes that provide useful sensory data when participants perform daily activities thus helping in HAR activity. Highly efficient features are extracted from this sensor data and techniques like denoising, normalization and segmentation are used to reduce noise and extract valuable feature vectors. Prior research showed that deep learning methods like recurrent neural networks and one-dimensional convolution networks provide excellent results in activity recognition tasks. In this paper, an ensemble model of CNN and SVM is proposed to further improve the accuracy and provide a robust model. Experimental methods are tested on UCI-HAR dataset and compared with other state-of-the-art methods like LSTM, CNN-LSTM, and Conv LSTM.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117092600","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936547
Baboo Barik, D. Srinivasan, K. Arulvendhan, Suresh N
A solar cell turns photon energy into electrical potential in a P-N junction (P-Type and N-Type), which are both equivalent circuits. While synchronizing with various grid and non-linear loads, the PV Photovoltaic input source comprises oscillations distorting, voltage sags/swell, and dc voltage of power quality concerns. The proposed technique for resolving the problem is Grid-connected output-based Photovoltaic (P.V.) System Power Quality Improvement. Proportional Integral (PI) Controllers are used in this method to control parameters like sampling rate and Improved Disrupt and Observe values, which have a substantial impact on the inter oscillatory form property of PV systems. The High gain (Step-Up) DC-DC Converter coupled based capacitor is recovered by the passive clamped circuit, which also limits the switch. Maximum power point tracking is a controller technique that provides inter harmonic emission, which is one of the most significant pieces of enhancing source voltage and current. The end result is improved power quality and gain without even any distortion in the Renewable Energy System's output.
{"title":"High step-up DC-DC Converter based Renewable Energy System for Improving Power Quality and Low Voltage Stress using PI Controller Technique","authors":"Baboo Barik, D. Srinivasan, K. Arulvendhan, Suresh N","doi":"10.1109/ICECAA55415.2022.9936547","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936547","url":null,"abstract":"A solar cell turns photon energy into electrical potential in a P-N junction (P-Type and N-Type), which are both equivalent circuits. While synchronizing with various grid and non-linear loads, the PV Photovoltaic input source comprises oscillations distorting, voltage sags/swell, and dc voltage of power quality concerns. The proposed technique for resolving the problem is Grid-connected output-based Photovoltaic (P.V.) System Power Quality Improvement. Proportional Integral (PI) Controllers are used in this method to control parameters like sampling rate and Improved Disrupt and Observe values, which have a substantial impact on the inter oscillatory form property of PV systems. The High gain (Step-Up) DC-DC Converter coupled based capacitor is recovered by the passive clamped circuit, which also limits the switch. Maximum power point tracking is a controller technique that provides inter harmonic emission, which is one of the most significant pieces of enhancing source voltage and current. The end result is improved power quality and gain without even any distortion in the Renewable Energy System's output.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123403893","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936206
A. P, Avinash Sharma, S. R. Kawale, S. P. Diwan, Dankan Gowda V
Unlike the healthy cells in the breast tissue, cancerous breast cells are unwelcome and have strange properties. In both sexes, this will quickly expand and infiltrate adjacent tissue, leading to the formation of a tumour. Using the Intelligent-Breast Abnormality Detection (I-BAD) framework, many breast cancer parameters are evaluated in this article. It has already been shown that some indicators may be used for early detection of breast cancer. There is also discussion of the instruments and strategies that facilitate the monitoring of the selected breast health metrics. Classification methods that use machine learning to store and analyse data are also discussed. The suggested I-BAD framework’s process is then visually shown in clean drawings.
{"title":"Intelligent Breast Abnormality Framework for Detection and Evaluation of Breast Abnormal Parameters","authors":"A. P, Avinash Sharma, S. R. Kawale, S. P. Diwan, Dankan Gowda V","doi":"10.1109/ICECAA55415.2022.9936206","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936206","url":null,"abstract":"Unlike the healthy cells in the breast tissue, cancerous breast cells are unwelcome and have strange properties. In both sexes, this will quickly expand and infiltrate adjacent tissue, leading to the formation of a tumour. Using the Intelligent-Breast Abnormality Detection (I-BAD) framework, many breast cancer parameters are evaluated in this article. It has already been shown that some indicators may be used for early detection of breast cancer. There is also discussion of the instruments and strategies that facilitate the monitoring of the selected breast health metrics. Classification methods that use machine learning to store and analyse data are also discussed. The suggested I-BAD framework’s process is then visually shown in clean drawings.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123278040","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936560
Junaid Ahmed Mohammed Abdul, Santhosh Kumar Dhatrika, P. Kumar
A Wireless Sensor Network is an infrastructure-free wireless network that uses an ad-hoc deployment of a large number of wireless sensors to monitor system, physical, and environmental factors. Sensor node energy consumption is a major determinant of wireless sensor network longevity. The Distributed Energy-aware Fuzzy Logic-based routing algorithm (DEFL) proposed in this paper aims to strike a compromise between energy efficiency measures balance. For the shortest path calculation, our architecture captures the network state using relevant energy measurements and maps them to cost values. I also added a Redundant Packet Monitoring Algorithm to each sensor node as a recommended technique, which attaches temporary memory to each sensor node and checks it anytime the sensor node senses any data.
{"title":"A Technique to Improve the Lifetime of Heterogeneous Wireless Sensor Networks by Removing Redundant Packets","authors":"Junaid Ahmed Mohammed Abdul, Santhosh Kumar Dhatrika, P. Kumar","doi":"10.1109/ICECAA55415.2022.9936560","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936560","url":null,"abstract":"A Wireless Sensor Network is an infrastructure-free wireless network that uses an ad-hoc deployment of a large number of wireless sensors to monitor system, physical, and environmental factors. Sensor node energy consumption is a major determinant of wireless sensor network longevity. The Distributed Energy-aware Fuzzy Logic-based routing algorithm (DEFL) proposed in this paper aims to strike a compromise between energy efficiency measures balance. For the shortest path calculation, our architecture captures the network state using relevant energy measurements and maps them to cost values. I also added a Redundant Packet Monitoring Algorithm to each sensor node as a recommended technique, which attaches temporary memory to each sensor node and checks it anytime the sensor node senses any data.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121904674","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936053
Ascharya Soni, Anuraag Khare, P. S. Darshan Balaji, Sachin Verma, K. P. Asha Rani, S. Gowrishankar
It is crucial to comprehend how insect pest populations affect the subsequent yield or harvest since the ultimate goal of agriculture is to provide a sustained economic production of crop products. Using pesticides is the simplest technique to manage the pest infestation. However, using pesticides improperly or in excess can harm both people and animals as well as the plants. Machine learning algorithms and image processing techniques are widely used in agricultural research, and they have significant potential, particularly in plant protection, which ultimately leads to crop management. This paper highlights the detection of pests and their control measures. A smartphone camera will capture photographs of the pests through a mobile app built using the Flutter framework. The images are then analyzed in the app using various transfer learning based models for available pest identification kaggle dataset. The flutter framework offers the ability to monitor targets in real-time on a mobile device and aids in visualizing the detected pest by integrating augmented reality on to the app.
{"title":"Pest Identification and Control using Deep Learning and Augmented Reality","authors":"Ascharya Soni, Anuraag Khare, P. S. Darshan Balaji, Sachin Verma, K. P. Asha Rani, S. Gowrishankar","doi":"10.1109/ICECAA55415.2022.9936053","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936053","url":null,"abstract":"It is crucial to comprehend how insect pest populations affect the subsequent yield or harvest since the ultimate goal of agriculture is to provide a sustained economic production of crop products. Using pesticides is the simplest technique to manage the pest infestation. However, using pesticides improperly or in excess can harm both people and animals as well as the plants. Machine learning algorithms and image processing techniques are widely used in agricultural research, and they have significant potential, particularly in plant protection, which ultimately leads to crop management. This paper highlights the detection of pests and their control measures. A smartphone camera will capture photographs of the pests through a mobile app built using the Flutter framework. The images are then analyzed in the app using various transfer learning based models for available pest identification kaggle dataset. The flutter framework offers the ability to monitor targets in real-time on a mobile device and aids in visualizing the detected pest by integrating augmented reality on to the app.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"66 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114074938","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936568
D. Ganesh, K. Suresh, M. S. Kumar, K. Balaji, Sreedhar Burada
As a result of this new computer design, edge computing can process data rapidly and effectively near to the source, avoiding network resource and latency constraints. By shifting computing power to the network edge, edge computing decreases the load on cloud services centers while also reducing the time required for users to input data. Edge computing advantages for data-intensive services, in particular, could be obscured if access latency becomes a bottleneck. Edge computing raises a number of challenges, such as security concerns, data incompleteness, and a hefty up-front and ongoing expense. There is now a shift in the worldwide mobile communications sector toward 5G technology. This unprecedented attention to edge computing has come about because 5G is one of the primary entry technologies for large-scale deployment. Edge computing privacy has been a major concern since the technology’s inception, limiting its adoption and advancement. As the capabilities of edge computing have evolved, so have the security issues that have arisen as a result of these developments, as well as the increasing public demand for privacy protection. The lack of trust amongst IoT devices is exacerbated by the inherent security concerns and assaults that plague IoT edge devices. A cognitive trust management system is proposed to reduce this malicious activity by maintaining the confidence of an appliance & managing the service level belief & Quality of Service (QoS). Improved packet delivery ratio and jitter in cognitive trust management systems based on QoS parameters show promise for spotting potentially harmful edge nodes in computing networks at the edge.
{"title":"Improving Security in Edge Computing by using Cognitive Trust Management Model","authors":"D. Ganesh, K. Suresh, M. S. Kumar, K. Balaji, Sreedhar Burada","doi":"10.1109/ICECAA55415.2022.9936568","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936568","url":null,"abstract":"As a result of this new computer design, edge computing can process data rapidly and effectively near to the source, avoiding network resource and latency constraints. By shifting computing power to the network edge, edge computing decreases the load on cloud services centers while also reducing the time required for users to input data. Edge computing advantages for data-intensive services, in particular, could be obscured if access latency becomes a bottleneck. Edge computing raises a number of challenges, such as security concerns, data incompleteness, and a hefty up-front and ongoing expense. There is now a shift in the worldwide mobile communications sector toward 5G technology. This unprecedented attention to edge computing has come about because 5G is one of the primary entry technologies for large-scale deployment. Edge computing privacy has been a major concern since the technology’s inception, limiting its adoption and advancement. As the capabilities of edge computing have evolved, so have the security issues that have arisen as a result of these developments, as well as the increasing public demand for privacy protection. The lack of trust amongst IoT devices is exacerbated by the inherent security concerns and assaults that plague IoT edge devices. A cognitive trust management system is proposed to reduce this malicious activity by maintaining the confidence of an appliance & managing the service level belief & Quality of Service (QoS). Improved packet delivery ratio and jitter in cognitive trust management systems based on QoS parameters show promise for spotting potentially harmful edge nodes in computing networks at the edge.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"270 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122467909","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}
Pub Date : 2022-10-13DOI: 10.1109/ICECAA55415.2022.9936409
D. Mahalakshmi, S. Appavu alias Balamurugan, M. Chinnadurai, D. Vaishnavi
Data processing and analytics are wide spread study with profound applications. Data analytics deals with deriving or applying an algorithm to an application that work with dataset. The proposed work analyses the image data with optimization algorithm by using novel method of Fire-Fly (FF) algorithm, which is named as Densely Search Fire-Fly (DSFF) optimization algorithm. The Neural Network (NN) is applied to classify the optimized data. In this process, the optimized data refers to selective attributes from the raw data of image features. To test the performance of proposed optimization, the Gabor feature extraction method is used to fetch the features from raw image data. The Gabor method retrieves the pattern in various angle of projections. This produces 5 × 8 number of patterns to represent the image feature. From this feature attributes of whole image dataset, the optimization search for the best attributes by the reference of weight value is calculated from the particles of Fire-Fly. According to the best selection of attributes from the objective function, the neurons in a network that can segregate the different classes in the training dataset. The performance of the proposed FF algorithm are compared with the traditional optimization methods in the image classification application.
{"title":"A Novel Densely Search based Fire-Fly (DSFF) Optimization Algorithm for Image Classification Application","authors":"D. Mahalakshmi, S. Appavu alias Balamurugan, M. Chinnadurai, D. Vaishnavi","doi":"10.1109/ICECAA55415.2022.9936409","DOIUrl":"https://doi.org/10.1109/ICECAA55415.2022.9936409","url":null,"abstract":"Data processing and analytics are wide spread study with profound applications. Data analytics deals with deriving or applying an algorithm to an application that work with dataset. The proposed work analyses the image data with optimization algorithm by using novel method of Fire-Fly (FF) algorithm, which is named as Densely Search Fire-Fly (DSFF) optimization algorithm. The Neural Network (NN) is applied to classify the optimized data. In this process, the optimized data refers to selective attributes from the raw data of image features. To test the performance of proposed optimization, the Gabor feature extraction method is used to fetch the features from raw image data. The Gabor method retrieves the pattern in various angle of projections. This produces 5 × 8 number of patterns to represent the image feature. From this feature attributes of whole image dataset, the optimization search for the best attributes by the reference of weight value is calculated from the particles of Fire-Fly. According to the best selection of attributes from the objective function, the neurons in a network that can segregate the different classes in the training dataset. The performance of the proposed FF algorithm are compared with the traditional optimization methods in the image classification application.","PeriodicalId":273850,"journal":{"name":"2022 International Conference on Edge Computing and Applications (ICECAA)","volume":"77 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125234714","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：