Pub Date : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009588
G. V. S. S. Santosh, G. C. Kumar, G. Sandeep, G. A. E. S. Kumar
Traffic signs provide the necessary information and warn of possible dangers. Traffic sign recognition plays a crucial role in helping drivers understand signposts, obey traffic rules and develop automated driving systems. This research work has developed a convolutional neural network (CNN) model to classify the traffic signs displayed in the image into different categories, such as speed limits, prohibitions, left or right turns, child crossings, overtaking heavy vehicles, etc. The proposed system can recognize and classify 43 types of signs. The proposed model has achieved an accuracy of 98.81% on test data.
{"title":"German Traffic Sign Recognition Using Convolutional Neural Network","authors":"G. V. S. S. Santosh, G. C. Kumar, G. Sandeep, G. A. E. S. Kumar","doi":"10.1109/ICECA55336.2022.10009588","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009588","url":null,"abstract":"Traffic signs provide the necessary information and warn of possible dangers. Traffic sign recognition plays a crucial role in helping drivers understand signposts, obey traffic rules and develop automated driving systems. This research work has developed a convolutional neural network (CNN) model to classify the traffic signs displayed in the image into different categories, such as speed limits, prohibitions, left or right turns, child crossings, overtaking heavy vehicles, etc. The proposed system can recognize and classify 43 types of signs. The proposed model has achieved an accuracy of 98.81% on test data.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121371786","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009592
B. VeeraSekharReddy, Koppula Srinivas Rao, Neeraja Koppula
Various Natural Language Processing (NLP) applications rely on Named Entity Recognition (NER) to help them sift through mountains of unstructured text data and find the information they need. Named Entity Recognition (NER) is the process of assigning labels to words in a text so that they can be sorted into categories. These state-of-the-art models achieve improved results despite limited resources, making language models increasingly valuable in a variety of NLP tasks. The Conditional Random Field and Active Learning Procedure form the basis of a novel Approach to named entity recognition discussed in this article. Following is an algorithmic description of how the AL-CRF model operates: Initially the samples are clustered with K-Means. Samples are used to train the fundamental CRF classifier, which is done by performing stratified sampling on the generated clusters. The following phase involves starting the selection process based on entropy. The training set is expanded to include examples with the greatest entropy values. The CRF classifier is then trained again using with a new training set, and the procedure is repeated. The AL's learning and selection procedure is repeatedly done until the harmonic mean stabilises and model for NER is obtained. The primary benefit of our method is that it is both more efficient and requires less manually marked training samples. Because of this, the procedure may become more reliable and cost-efficient.
{"title":"Named Entity Recognition using CRF with Active Learning Algorithm in English Texts","authors":"B. VeeraSekharReddy, Koppula Srinivas Rao, Neeraja Koppula","doi":"10.1109/ICECA55336.2022.10009592","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009592","url":null,"abstract":"Various Natural Language Processing (NLP) applications rely on Named Entity Recognition (NER) to help them sift through mountains of unstructured text data and find the information they need. Named Entity Recognition (NER) is the process of assigning labels to words in a text so that they can be sorted into categories. These state-of-the-art models achieve improved results despite limited resources, making language models increasingly valuable in a variety of NLP tasks. The Conditional Random Field and Active Learning Procedure form the basis of a novel Approach to named entity recognition discussed in this article. Following is an algorithmic description of how the AL-CRF model operates: Initially the samples are clustered with K-Means. Samples are used to train the fundamental CRF classifier, which is done by performing stratified sampling on the generated clusters. The following phase involves starting the selection process based on entropy. The training set is expanded to include examples with the greatest entropy values. The CRF classifier is then trained again using with a new training set, and the procedure is repeated. The AL's learning and selection procedure is repeatedly done until the harmonic mean stabilises and model for NER is obtained. The primary benefit of our method is that it is both more efficient and requires less manually marked training samples. Because of this, the procedure may become more reliable and cost-efficient.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"144 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"113991175","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009184
I. Haritha, M. Harshini, Shruti Patil, Jeethu Philip
This research work aims to perform object detection by using the You Look Only Once (YOLO) method. This method is much efficient to the existing models in terms of speed and performance. Some of the algorithms do not scan all the regions in single forward propagation but in YOLO, the algorithm analyzes the entire image by predicting binding boxes using convolutional neural network and class opportunities. YOLO performs faster when compared to other algorithms.
本研究的目的是利用You Look Only Once (YOLO)方法进行目标检测。该方法在速度和性能上都比现有模型有效。一些算法在单次前向传播中没有扫描所有区域，但在YOLO中，算法通过使用卷积神经网络和类机会预测绑定框来分析整个图像。与其他算法相比，YOLO的执行速度更快。
{"title":"Real Time Object Detection using YOLO Algorithm","authors":"I. Haritha, M. Harshini, Shruti Patil, Jeethu Philip","doi":"10.1109/ICECA55336.2022.10009184","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009184","url":null,"abstract":"This research work aims to perform object detection by using the You Look Only Once (YOLO) method. This method is much efficient to the existing models in terms of speed and performance. Some of the algorithms do not scan all the regions in single forward propagation but in YOLO, the algorithm analyzes the entire image by predicting binding boxes using convolutional neural network and class opportunities. YOLO performs faster when compared to other algorithms.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124075624","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009207
G. Deepanraj, L. Kalaivani
Overheating in high-voltage equipment is detrimental to its reliability. Insulated equipment such as bushings plays a predominant role in transformer applications. In bushing, thermal characteristics are a key factor, and they act significantly in various conditions. During abnormal conditions, it experiences thermal stress due to dielectric loss, fault current, natural disasters, etc. This paper emphasizes the idea of designing the thermal model of porcelain bushing, analysing the temperature site, and then overcoming the negative impact of the HV bushing. Finding the bushing's maximum low temperature location and analyzing solutions to this issue are the papers goals. Stationary and time-dependent effects were studied using the advanced finite element method (AFEM). The proposed heat transfer model is examined at 11 kV, 273A in an 11 kV porcelain bushing. To the suggested thermal model's accuracy or predicted reading as well as the parameter responsible for the temperature increase are the problems of this work.
{"title":"Computational Model for Transformer Bushing using Advanced Finite Element Method","authors":"G. Deepanraj, L. Kalaivani","doi":"10.1109/ICECA55336.2022.10009207","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009207","url":null,"abstract":"Overheating in high-voltage equipment is detrimental to its reliability. Insulated equipment such as bushings plays a predominant role in transformer applications. In bushing, thermal characteristics are a key factor, and they act significantly in various conditions. During abnormal conditions, it experiences thermal stress due to dielectric loss, fault current, natural disasters, etc. This paper emphasizes the idea of designing the thermal model of porcelain bushing, analysing the temperature site, and then overcoming the negative impact of the HV bushing. Finding the bushing's maximum low temperature location and analyzing solutions to this issue are the papers goals. Stationary and time-dependent effects were studied using the advanced finite element method (AFEM). The proposed heat transfer model is examined at 11 kV, 273A in an 11 kV porcelain bushing. To the suggested thermal model's accuracy or predicted reading as well as the parameter responsible for the temperature increase are the problems of this work.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128085703","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009402
S. V, G. Kiran, Yashwanth Guntupalli, Ch Navya Gayathri, A. Raju
Despite their great accuracy, neural networks are not very popular in fields like medical, finance, education, and others where predictive explainability are essential. The objective of this work is to create and train a model using PyTorch Pipeline that divides photos into “Good” and “Anomaly” classes and, if the image is categorized as an “Anomaly,” a bounding box is returned for the fault. While this work appears straightforward and similar to other item detection tasks, there is a problem that it lacks bounding box labels. Fortunately, this problem can be solved by the model in the inference mode, trained without labels for defective regions, and is able to forecast a bounding box for a defective region in the picture, by processing feature maps from the deep convolutional layers. This work discusses the strategy and talks about how to use it for the purpose of defect detection in the real world. A 400-image dataset that includes pictures of both perfect objects (classed as “good”) and imperfect objects (classed as “anomalies”) has been used. The dataset is unbalanced; there are more examples of good than bad photographs. Any form of object, such as a bottle, cable, pill, tile, piece of leather, a zipper, etc., may be seen in the images.
{"title":"Automatic Visual Inspection - Defects Detection using CNN","authors":"S. V, G. Kiran, Yashwanth Guntupalli, Ch Navya Gayathri, A. Raju","doi":"10.1109/ICECA55336.2022.10009402","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009402","url":null,"abstract":"Despite their great accuracy, neural networks are not very popular in fields like medical, finance, education, and others where predictive explainability are essential. The objective of this work is to create and train a model using PyTorch Pipeline that divides photos into “Good” and “Anomaly” classes and, if the image is categorized as an “Anomaly,” a bounding box is returned for the fault. While this work appears straightforward and similar to other item detection tasks, there is a problem that it lacks bounding box labels. Fortunately, this problem can be solved by the model in the inference mode, trained without labels for defective regions, and is able to forecast a bounding box for a defective region in the picture, by processing feature maps from the deep convolutional layers. This work discusses the strategy and talks about how to use it for the purpose of defect detection in the real world. A 400-image dataset that includes pictures of both perfect objects (classed as “good”) and imperfect objects (classed as “anomalies”) has been used. The dataset is unbalanced; there are more examples of good than bad photographs. Any form of object, such as a bottle, cable, pill, tile, piece of leather, a zipper, etc., may be seen in the images.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"125 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125704027","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009540
Ahsan Kabir Nuhel, Mir Mohibullah Sazid, Md. Nafim Mahmud Bhuiyan, Ariful Islam Arif, Priyadarshini Hriddhi Roy, Md Riazul Islam
A fully automated driving system allows an autonomous vehicle to adapt to external conditions that a human driver would typically handle. Using deep learning (DL), machine learning (ML), computer vision (CV), and conventional neural networks (CNN), a self-driving automobile will be developed in this study by modeling the design of a car body and the implementation of various sensors on the car chassis in order to automatically run the cars. In addition, the car will undergo real-time obstetrical on the roads and self-training in order to learn itself appropriately. In contrast, the fundamental principles of artificial intelligence and their relationship to the autonomous car are examined.
{"title":"Developing a Self-Driving Autonomous Car using Artificial Intelligence Algorithm","authors":"Ahsan Kabir Nuhel, Mir Mohibullah Sazid, Md. Nafim Mahmud Bhuiyan, Ariful Islam Arif, Priyadarshini Hriddhi Roy, Md Riazul Islam","doi":"10.1109/ICECA55336.2022.10009540","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009540","url":null,"abstract":"A fully automated driving system allows an autonomous vehicle to adapt to external conditions that a human driver would typically handle. Using deep learning (DL), machine learning (ML), computer vision (CV), and conventional neural networks (CNN), a self-driving automobile will be developed in this study by modeling the design of a car body and the implementation of various sensors on the car chassis in order to automatically run the cars. In addition, the car will undergo real-time obstetrical on the roads and self-training in order to learn itself appropriately. In contrast, the fundamental principles of artificial intelligence and their relationship to the autonomous car are examined.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132628423","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009075
N. Velmurugan, Christika. S, Govarthini. V, K. S.
A Sanction Probe program proposed in this paper is primarily concerned with the uncertainty that arises while checking the hosteller's authorization to depart. A program that uses image processing has been developed to automatically check a student's permission as they walk through the college's entrance. In a hostel that has many students, it is difficult to maintain a manual outing record. Students intend to go on an outing or to their native during festivals/ government holidays. The developed Sanction Probe software is a gadget that checks a student's authorization before they leave the dormitory. Hostel guests are required to submit their authorization in an online form through the Thumbs-up website to the hostel's website under their username. When a hosteller walks through the entry, their accounts will be verified to determine whether they have an authorization, and a sound will alert if not. Machine learning is used to create the software by training sample inputs and training the model for image recognition.
{"title":"Thumbs-Up: A Sanction Probe Software using Machine Learning","authors":"N. Velmurugan, Christika. S, Govarthini. V, K. S.","doi":"10.1109/ICECA55336.2022.10009075","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009075","url":null,"abstract":"A Sanction Probe program proposed in this paper is primarily concerned with the uncertainty that arises while checking the hosteller's authorization to depart. A program that uses image processing has been developed to automatically check a student's permission as they walk through the college's entrance. In a hostel that has many students, it is difficult to maintain a manual outing record. Students intend to go on an outing or to their native during festivals/ government holidays. The developed Sanction Probe software is a gadget that checks a student's authorization before they leave the dormitory. Hostel guests are required to submit their authorization in an online form through the Thumbs-up website to the hostel's website under their username. When a hosteller walks through the entry, their accounts will be verified to determine whether they have an authorization, and a sound will alert if not. Machine learning is used to create the software by training sample inputs and training the model for image recognition.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133802847","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009277
Hruthvik Naik, Kakumanu Yashwanth, S. P, N. Jayapandian
Sales forecasting is crucial in the food industry, which experiences high levels of food sales and demand. The industry has concentrated on a well-known and established statistical model. Due to modern technologies, it has gained tremendous appeal in improving market operations and productivity. The main objective is to find the most accurate algorithms to predict food sales and which algorithm is most suitable for sales forecasting. This research work has mentioned and discussed about several research articles that revolve around the techniques usedfor sales prediction as well as finding out the advantages and disadvantages of the said techniques. Various techniques were discussed as to predicting the sales but mainly Incline Increasing Regression and Accidental Forestry Lapse is used for attention. The manufacturing has concentrated on a well-known and established statistical model. Although algorithms like Modest Direct Regression, Incline Increasing Lapse, Provision Course Lapse, Accidental Forest Lapse, Gradient Boosting Regression, and Random Forest Regression are well familiar for outdoing others, it has remained decisively established that Random Forest Regression is the most appropriate technique when associated to the others. After doing the whole examination, the Random Forest Regression technique fared well when compared to other algorithms. The feature importance is generated for the selected dataset using Python and Random Forest Regression and the nose position chart is also explainedin detail. The proposed model is compared three major parameters that are accuracy score, mean absolute error and max error. The proposed random forest regression accuracy score is improved nearly 1.83% and absolute error rate is reduced 4.66%.
{"title":"Machine Learning based Food Sales Prediction using Random Forest Regression","authors":"Hruthvik Naik, Kakumanu Yashwanth, S. P, N. Jayapandian","doi":"10.1109/ICECA55336.2022.10009277","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009277","url":null,"abstract":"Sales forecasting is crucial in the food industry, which experiences high levels of food sales and demand. The industry has concentrated on a well-known and established statistical model. Due to modern technologies, it has gained tremendous appeal in improving market operations and productivity. The main objective is to find the most accurate algorithms to predict food sales and which algorithm is most suitable for sales forecasting. This research work has mentioned and discussed about several research articles that revolve around the techniques usedfor sales prediction as well as finding out the advantages and disadvantages of the said techniques. Various techniques were discussed as to predicting the sales but mainly Incline Increasing Regression and Accidental Forestry Lapse is used for attention. The manufacturing has concentrated on a well-known and established statistical model. Although algorithms like Modest Direct Regression, Incline Increasing Lapse, Provision Course Lapse, Accidental Forest Lapse, Gradient Boosting Regression, and Random Forest Regression are well familiar for outdoing others, it has remained decisively established that Random Forest Regression is the most appropriate technique when associated to the others. After doing the whole examination, the Random Forest Regression technique fared well when compared to other algorithms. The feature importance is generated for the selected dataset using Python and Random Forest Regression and the nose position chart is also explainedin detail. The proposed model is compared three major parameters that are accuracy score, mean absolute error and max error. The proposed random forest regression accuracy score is improved nearly 1.83% and absolute error rate is reduced 4.66%.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115346346","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009632
Mihir Prajapati, Mitul Nakrani, Tarjni Vyas, Lata Gohil, Shivani Desai, S. Degadwala
Stack Overflow is a well-known website which is utilized by nearly everyone who learns to code, share their knowledge and publicly participate in this question-answering forum. The questions posted on the Stack Overflow forum by a user requires a minimum of 1 tag to be manually entered in by them. Tagging most commonly means to associate some single word information about the context of given text or question. Tagging a question is useful in identifying the category that a question or text belongs. It is also beneficial in providing ease of access to a person having a requirement of specific categories of questions. On analysis of tags associated with the questions on the website, it was found that a large number of the questions are labelled by more than one tags, with many of them not being tagged accurately. Due to this situation, it becomes challenging for the users to search for relevant tags. So, the main aim of this research task is to explore methods and compare different techniques in order to create an auto tagging system with the aid of Machine learning and deep learning facilities, accompanied by data preprocessing steps. The dataset for this purpose was taken from Kaggle, known as StackSample dataset, which is a dataset containing 10 percent of the questions present on the website. The output of the research performed for this purpose provided satisfactory results with scope of improvement.
{"title":"Automatic Question Tagging using Machine Learning and Deep learning Algorithms","authors":"Mihir Prajapati, Mitul Nakrani, Tarjni Vyas, Lata Gohil, Shivani Desai, S. Degadwala","doi":"10.1109/ICECA55336.2022.10009632","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009632","url":null,"abstract":"Stack Overflow is a well-known website which is utilized by nearly everyone who learns to code, share their knowledge and publicly participate in this question-answering forum. The questions posted on the Stack Overflow forum by a user requires a minimum of 1 tag to be manually entered in by them. Tagging most commonly means to associate some single word information about the context of given text or question. Tagging a question is useful in identifying the category that a question or text belongs. It is also beneficial in providing ease of access to a person having a requirement of specific categories of questions. On analysis of tags associated with the questions on the website, it was found that a large number of the questions are labelled by more than one tags, with many of them not being tagged accurately. Due to this situation, it becomes challenging for the users to search for relevant tags. So, the main aim of this research task is to explore methods and compare different techniques in order to create an auto tagging system with the aid of Machine learning and deep learning facilities, accompanied by data preprocessing steps. The dataset for this purpose was taken from Kaggle, known as StackSample dataset, which is a dataset containing 10 percent of the questions present on the website. The output of the research performed for this purpose provided satisfactory results with scope of improvement.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115469135","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 : 2022-12-01DOI: 10.1109/ICECA55336.2022.10009130
V. Surya, M. Selvam
Intrusion detection system to secure loT is a sophisticated tool to detect possible intrusions in the network and ensures confidentiality, integrity, and availability. loT is a precious domain that improves the standard of life, which cannot be accomplished in the existing conventional paradigm. The intrusion detection system is effective in identifying whether the attack is normal or not. Thus, classification algorithms can be applied for prediction. The Machine Learning and Deep Learning concepts of AI technology which contribute more to Data Science have produced remarkable developments in loT applications. In this paper, Machine Learning (ML) algorithms are used to secure loT devices using intrusion detection systems while working on loTID20 dataset. This dataset is highly imbalanced and contains different types of attacks and sub-attacks. The effect of the oversampling technique, Synthetic Minority Oversampling Technique (S MOTE) to balance the dataset significantly, has influenced the result. loT ID20 is a supervised dataset and different classification algorithms are used to measure the performance metrics namely, Accuracy, Recall, Precision, and F-score. The Binary and Multi classifications are done on the dataset using ML techniques. It is found that the accuracy obtained using the ML classifiers such as K-N earest Neighbor, Decision tree and Random Forest techniques is above 90%, showing that the mitigation of attacks that occur on an loT network is effective.
{"title":"An Effective Machine Learning Approach for loT Intrusion Detection System based on SMOTE","authors":"V. Surya, M. Selvam","doi":"10.1109/ICECA55336.2022.10009130","DOIUrl":"https://doi.org/10.1109/ICECA55336.2022.10009130","url":null,"abstract":"Intrusion detection system to secure loT is a sophisticated tool to detect possible intrusions in the network and ensures confidentiality, integrity, and availability. loT is a precious domain that improves the standard of life, which cannot be accomplished in the existing conventional paradigm. The intrusion detection system is effective in identifying whether the attack is normal or not. Thus, classification algorithms can be applied for prediction. The Machine Learning and Deep Learning concepts of AI technology which contribute more to Data Science have produced remarkable developments in loT applications. In this paper, Machine Learning (ML) algorithms are used to secure loT devices using intrusion detection systems while working on loTID20 dataset. This dataset is highly imbalanced and contains different types of attacks and sub-attacks. The effect of the oversampling technique, Synthetic Minority Oversampling Technique (S MOTE) to balance the dataset significantly, has influenced the result. loT ID20 is a supervised dataset and different classification algorithms are used to measure the performance metrics namely, Accuracy, Recall, Precision, and F-score. The Binary and Multi classifications are done on the dataset using ML techniques. It is found that the accuracy obtained using the ML classifiers such as K-N earest Neighbor, Decision tree and Random Forest techniques is above 90%, showing that the mitigation of attacks that occur on an loT network is effective.","PeriodicalId":356949,"journal":{"name":"2022 6th International Conference on Electronics, Communication and Aerospace Technology","volume":"275 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115886703","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：