Pub Date : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544902
S. Gayathri, T. Ujwala, C. Vinusha, N. Pauline, D.B. Tharunika
Papaya is a berry fruit with nutritional as well as real worth because to its non-seasonality and short harvesting period. The fiscal year 2020 statistics shows that the volume of papaya production increased to over six million metric tons in India. The grading of papayas is done by hand by human operators, which might lead to misclassifications. The identification of the ripeness of a fruit is important in case of distributing the classified papaya packages as well as in purchasing them by customers. Many projects were proposed earlier for classifying fruits and vegetables, however they were done using machine learning algorithms while the proposed system focuses on deep learning algorithm, especially using Convolution Neural Network (CNN). Convolution Neural Network is a deep learning technique that identifies features without the need for manual absorption. The papaya dataset which is used for this system consist of 300 images, in which each class (ripe, unripe and partially ripe) has 100 images. The proposed model is expected to have a maximum accuracy.
{"title":"Detection of Papaya Ripeness Using Deep Learning Approach","authors":"S. Gayathri, T. Ujwala, C. Vinusha, N. Pauline, D.B. Tharunika","doi":"10.1109/ICIRCA51532.2021.9544902","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544902","url":null,"abstract":"Papaya is a berry fruit with nutritional as well as real worth because to its non-seasonality and short harvesting period. The fiscal year 2020 statistics shows that the volume of papaya production increased to over six million metric tons in India. The grading of papayas is done by hand by human operators, which might lead to misclassifications. The identification of the ripeness of a fruit is important in case of distributing the classified papaya packages as well as in purchasing them by customers. Many projects were proposed earlier for classifying fruits and vegetables, however they were done using machine learning algorithms while the proposed system focuses on deep learning algorithm, especially using Convolution Neural Network (CNN). Convolution Neural Network is a deep learning technique that identifies features without the need for manual absorption. The papaya dataset which is used for this system consist of 300 images, in which each class (ripe, unripe and partially ripe) has 100 images. The proposed model is expected to have a maximum accuracy.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127254514","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544962
Ifrah Raoof, M. Gupta
Brain-computer interface provides an alternative way to communicate between the human brain and the external devices. Deep learning approaches have been widely used in various fields for feature extraction and classification task. However, the deep learning method requires a lot of data for training purpose. Due to the hectic calibration process, it is very difficult to collect large amount of EEG data. In such situations, deep neural network has proven very challenging in practice. This paper provides a comprehensive review of the various semi supervised approaches that have been used till now for the augmentation of motor imagery EEG data. Further, this research work has discussed about various research challenges faced by this field.
{"title":"A Study of Semi Supervised based approaches for Motor Imagery Signal Generation","authors":"Ifrah Raoof, M. Gupta","doi":"10.1109/ICIRCA51532.2021.9544962","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544962","url":null,"abstract":"Brain-computer interface provides an alternative way to communicate between the human brain and the external devices. Deep learning approaches have been widely used in various fields for feature extraction and classification task. However, the deep learning method requires a lot of data for training purpose. Due to the hectic calibration process, it is very difficult to collect large amount of EEG data. In such situations, deep neural network has proven very challenging in practice. This paper provides a comprehensive review of the various semi supervised approaches that have been used till now for the augmentation of motor imagery EEG data. Further, this research work has discussed about various research challenges faced by this field.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127509736","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544861
Uppuluri Ruchitha Venkata Sai Meenakshi, V. Jindal
Chronic Obstructive Pulmonary Disease (COPD) is characterized by a chronic airflow limitation that is generally progressive and an increased chronic inflammatory response triggered by harmful particles or gases in the airways. In general, symptoms, medical history, clinical examination, and lung ventilation obstruction play a vital role in diagnosis. However, COPD is treatable, even though it is a chronic condition that worsens over time. Furthermore, most patients with COPD can have improved symptom regulation and quality of life with careful treatment and a lower chance of developing other disorders. Therefore, COPD diagnosis is essential in the early stages, as it is treatable and will significantly impact the recovery of a patient's health. With tens of thousands of characteristics in high-dimensional biomedical data, precise and effective identification of the main characteristics in these data might help identify associated disorders. However, biological data frequently contains many irrelevant or duplicated characteristics, which significantly impact later classification accuracy and machine learning efficiency. As a result, for COPD diagnosis, an effective predictive model is needed. This study proposed a hybrid feature selection model to extract the best features from the high-dimensional data. These features are further passed to the classification models to identify the performance of the features on various classification models. According to the experimental data, the suggested hybrid feature selection model could predict COPD with a 95.18 percent accuracy and a Kappa Statistic of 0.9.
{"title":"A Hybrid Feature Selection Model for Predicting Chronic Obstructive Pulmonary Disease","authors":"Uppuluri Ruchitha Venkata Sai Meenakshi, V. Jindal","doi":"10.1109/ICIRCA51532.2021.9544861","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544861","url":null,"abstract":"Chronic Obstructive Pulmonary Disease (COPD) is characterized by a chronic airflow limitation that is generally progressive and an increased chronic inflammatory response triggered by harmful particles or gases in the airways. In general, symptoms, medical history, clinical examination, and lung ventilation obstruction play a vital role in diagnosis. However, COPD is treatable, even though it is a chronic condition that worsens over time. Furthermore, most patients with COPD can have improved symptom regulation and quality of life with careful treatment and a lower chance of developing other disorders. Therefore, COPD diagnosis is essential in the early stages, as it is treatable and will significantly impact the recovery of a patient's health. With tens of thousands of characteristics in high-dimensional biomedical data, precise and effective identification of the main characteristics in these data might help identify associated disorders. However, biological data frequently contains many irrelevant or duplicated characteristics, which significantly impact later classification accuracy and machine learning efficiency. As a result, for COPD diagnosis, an effective predictive model is needed. This study proposed a hybrid feature selection model to extract the best features from the high-dimensional data. These features are further passed to the classification models to identify the performance of the features on various classification models. According to the experimental data, the suggested hybrid feature selection model could predict COPD with a 95.18 percent accuracy and a Kappa Statistic of 0.9.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"220 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125158655","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}
Multimedia is considered as one of the best sources of entertainment. People of all age groups love to watch movies. Movie Recommender System is essential in our social lives as it enhances the field of entertainment. The proposed system on Movie Recommendation System caters the requirements of the user. The major aim is to provide crisp relevant content to the end-users out of semi-structured content on the internet. The main purpose is to generate accurate, efficient and personalized recommendations to the user. Various building blocks of the paper like Introduction, Literature Survey, Proposed System, Implementation & Result, Comparative Analysis, Conclusion and Future Work are discussed in detail. The proposed machine learning model is trained, tested, and a sentiment classifier is generated which classify the sentiments as a good or a bad sentiment. The recommender system is generated by applying Cosine similarity and making API Calls. As a result, the live working of the system generates accurate and personalized recommendations along with the analysis of sentiments for the end users. It is also concluded that Cosine Similarity provides better and efficient results for a recommender system.
{"title":"Movie Recommendation System using Cosine Similarity with Sentiment Analysis","authors":"Harsh Khatter, Nishtha Goel, Naina Gupta, Muskan Gulati","doi":"10.1109/ICIRCA51532.2021.9544794","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544794","url":null,"abstract":"Multimedia is considered as one of the best sources of entertainment. People of all age groups love to watch movies. Movie Recommender System is essential in our social lives as it enhances the field of entertainment. The proposed system on Movie Recommendation System caters the requirements of the user. The major aim is to provide crisp relevant content to the end-users out of semi-structured content on the internet. The main purpose is to generate accurate, efficient and personalized recommendations to the user. Various building blocks of the paper like Introduction, Literature Survey, Proposed System, Implementation & Result, Comparative Analysis, Conclusion and Future Work are discussed in detail. The proposed machine learning model is trained, tested, and a sentiment classifier is generated which classify the sentiments as a good or a bad sentiment. The recommender system is generated by applying Cosine similarity and making API Calls. As a result, the live working of the system generates accurate and personalized recommendations along with the analysis of sentiments for the end users. It is also concluded that Cosine Similarity provides better and efficient results for a recommender system.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124745544","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544925
Aarushi Dua, A. Bhatia, B. Kalra, Srishti Vashishtha
This paper presents a way for generating online handwriting using voice. To build this tool, two broad steps are required: Voice Recognition using Google Speech-to-text API and Handwritten Recognition using a combination of Recurrent and Convolutional neural networks (RCNN). The model is evaluated on IAM and Electronic Fonts datasets that contains handwritten images. This research work has reported the result of training data based on Connectionist Temporal Classification (CTC) loss. CTC also has a function named decoder to predict vector data generated by RCNN into understandable text.
{"title":"A Novel Recurrent and Convolutional Neural Network Technique for Generating Handwriting from Voice","authors":"Aarushi Dua, A. Bhatia, B. Kalra, Srishti Vashishtha","doi":"10.1109/ICIRCA51532.2021.9544925","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544925","url":null,"abstract":"This paper presents a way for generating online handwriting using voice. To build this tool, two broad steps are required: Voice Recognition using Google Speech-to-text API and Handwritten Recognition using a combination of Recurrent and Convolutional neural networks (RCNN). The model is evaluated on IAM and Electronic Fonts datasets that contains handwritten images. This research work has reported the result of training data based on Connectionist Temporal Classification (CTC) loss. CTC also has a function named decoder to predict vector data generated by RCNN into understandable text.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122070907","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544604
M. Ramkumar, G. Swapna, A. Saravanan, N. Hemalatha, G. Dharmaraj, S. Purushotham, M. Sivaramkrishnan.
Due to the ever-increasing concern for the environment and the progression of technology, renewable energy such as solar photovoltaic (PV), wind, and super capacitor is being widely used. Many creative approaches have been used to convert the power from renewable sources. One such creative solution is using power electronic converters to match the load and grid requirements so that the renewable generation's dynamic but instead steady-state characteristics are enhanced, with the goal of achieving maximum power point tracking (MPPT) regulate and energy storage to resolve this issue. This new design seeks to increase circuit efficacy and power density by using a multiple DC-DC converter [3] which has a DC input port for renewable sources, an unidirectional Input voltage port for energy storage, as well as an Output signal port for operating the load. A few new DC-DC four converters have developed in recent years and are now being researched in the literature. This study reviews several three-port DC/DC converter topologies that have been developed by different research organizations. The study concludes that topologies based on three-port Power converter with power terminals and a single inductor are likely for further research. The suggested system's simulation is done using MATLAB/SIMULINK.
{"title":"Super Capacitor based Solar and Wind Grid Connected Storage System","authors":"M. Ramkumar, G. Swapna, A. Saravanan, N. Hemalatha, G. Dharmaraj, S. Purushotham, M. Sivaramkrishnan.","doi":"10.1109/ICIRCA51532.2021.9544604","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544604","url":null,"abstract":"Due to the ever-increasing concern for the environment and the progression of technology, renewable energy such as solar photovoltaic (PV), wind, and super capacitor is being widely used. Many creative approaches have been used to convert the power from renewable sources. One such creative solution is using power electronic converters to match the load and grid requirements so that the renewable generation's dynamic but instead steady-state characteristics are enhanced, with the goal of achieving maximum power point tracking (MPPT) regulate and energy storage to resolve this issue. This new design seeks to increase circuit efficacy and power density by using a multiple DC-DC converter [3] which has a DC input port for renewable sources, an unidirectional Input voltage port for energy storage, as well as an Output signal port for operating the load. A few new DC-DC four converters have developed in recent years and are now being researched in the literature. This study reviews several three-port DC/DC converter topologies that have been developed by different research organizations. The study concludes that topologies based on three-port Power converter with power terminals and a single inductor are likely for further research. The suggested system's simulation is done using MATLAB/SIMULINK.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"187 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123161440","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544534
Shalini Kumari, Sandeep Singh Kang
The hidden node issue is a well-known phenomenon in IEEE 802.11 wireless networks. This research work show that the well-known ready-to-send / clear-to-send (RTS / CTS) approach, which is used to solve the hidden node problem, is ineffective in this case. We conducted real-world network experiments to examine the impact of hidden nodes in infrastructure as well as ad hoc multi-hop networks. Transmission and Carrier sensing channel models are proposed in this investigation. As a solution to the hidden node problem, this research work will also study the RTS / CTS mode. The proposed model utilizes 2 Mbps or 11 Mbps to transmit RTS / CTS not only solve the problem but also degrades the performance by introducing additional over ad network. This paper attempts to identify the basic conditions that lead to the hidden node. In particular, the proposed research work shows that the occurrence of hidden node is primarily due to the limitations of the 802.11 protocol. Based on the insight gained from the study, this research work is designing a hidden-node-free model that eliminates the hidden node entirely.
{"title":"Hidden Node Problem in Remote Ad-Hoc Networks","authors":"Shalini Kumari, Sandeep Singh Kang","doi":"10.1109/ICIRCA51532.2021.9544534","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544534","url":null,"abstract":"The hidden node issue is a well-known phenomenon in IEEE 802.11 wireless networks. This research work show that the well-known ready-to-send / clear-to-send (RTS / CTS) approach, which is used to solve the hidden node problem, is ineffective in this case. We conducted real-world network experiments to examine the impact of hidden nodes in infrastructure as well as ad hoc multi-hop networks. Transmission and Carrier sensing channel models are proposed in this investigation. As a solution to the hidden node problem, this research work will also study the RTS / CTS mode. The proposed model utilizes 2 Mbps or 11 Mbps to transmit RTS / CTS not only solve the problem but also degrades the performance by introducing additional over ad network. This paper attempts to identify the basic conditions that lead to the hidden node. In particular, the proposed research work shows that the occurrence of hidden node is primarily due to the limitations of the 802.11 protocol. Based on the insight gained from the study, this research work is designing a hidden-node-free model that eliminates the hidden node entirely.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131859796","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544966
B. T, T. Sujatha, S. Premnath, V. Devi, A. Benő, S. S. C. Mary
The Fuzzy Agent Computing System is a competitive way of establishing an interactive middleware component in a Ubiquitous Computing Environment (UCE). However, there are some difficulties faced along the way such as high component building time imposed on users working in a heterogeneous environment and also high memory consumption. To make the middleware adapt to the users benefit, the proposed fuzzy agent computing system attempts to work in an online deep-rooted learning methodology. The purpose of this work is to establish a full-fledged connection between the data innovation gear and the individuals with the help of UCE devices in an undetectable network. It ensures that users prerequisites are fulfilled with this dynamically built computational environment. Because of the vast database available online without metadata repository and ontology, finding the apt service that will meet customers' requirements, proves to be a hassle. To aid the end users with the necessary services, a fuzzy agent computing system in an ubiquitous computing environment is proposed in this work resulting in reduced CBT and MC. This work focuses on the communication between the device and the user to create quick access to the administration and elements available in the Ubiquitous Computing Environment.
{"title":"Design of an Efficient User Interface for Ubiquitous Soft Computing Environment","authors":"B. T, T. Sujatha, S. Premnath, V. Devi, A. Benő, S. S. C. Mary","doi":"10.1109/ICIRCA51532.2021.9544966","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544966","url":null,"abstract":"The Fuzzy Agent Computing System is a competitive way of establishing an interactive middleware component in a Ubiquitous Computing Environment (UCE). However, there are some difficulties faced along the way such as high component building time imposed on users working in a heterogeneous environment and also high memory consumption. To make the middleware adapt to the users benefit, the proposed fuzzy agent computing system attempts to work in an online deep-rooted learning methodology. The purpose of this work is to establish a full-fledged connection between the data innovation gear and the individuals with the help of UCE devices in an undetectable network. It ensures that users prerequisites are fulfilled with this dynamically built computational environment. Because of the vast database available online without metadata repository and ontology, finding the apt service that will meet customers' requirements, proves to be a hassle. To aid the end users with the necessary services, a fuzzy agent computing system in an ubiquitous computing environment is proposed in this work resulting in reduced CBT and MC. This work focuses on the communication between the device and the user to create quick access to the administration and elements available in the Ubiquitous Computing Environment.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130559799","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544592
Anirudh Ghildiyal, Sachin Sharma
The rise of music industry across the globe can be seen with the new type of genre being created, and more artist and musicians joining this profession. A lot of music is created and launched every day. A major task for various music streaming platform is to classify these songs based on the genres and recommend music to the users. To overcome this many artificial intelligence algorithms are developed. One of the major problems in designing machine learning models is inadequate data for training. Certain dataset contains lot of redundant features that could cause the models to overfit. This problem could be resolved by data filtering. This paper has developed the multiple Artificial Intelligence (AI) models and applied data filtering method on the GTZAN dataset for music genre classification. A comparative analysis is done and discussed in this paper.
{"title":"Music Genre Classification Using Data Filtering Algorithm: An Artificial Intelligence Approach","authors":"Anirudh Ghildiyal, Sachin Sharma","doi":"10.1109/ICIRCA51532.2021.9544592","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544592","url":null,"abstract":"The rise of music industry across the globe can be seen with the new type of genre being created, and more artist and musicians joining this profession. A lot of music is created and launched every day. A major task for various music streaming platform is to classify these songs based on the genres and recommend music to the users. To overcome this many artificial intelligence algorithms are developed. One of the major problems in designing machine learning models is inadequate data for training. Certain dataset contains lot of redundant features that could cause the models to overfit. This problem could be resolved by data filtering. This paper has developed the multiple Artificial Intelligence (AI) models and applied data filtering method on the GTZAN dataset for music genre classification. A comparative analysis is done and discussed in this paper.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124684618","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 : 2021-09-02DOI: 10.1109/ICIRCA51532.2021.9544889
N. Janapriya, K. Anuradha, V. Srilakshmi
Adversarial machine learning calculations handle adversarial instance age, producing bogus data information with the ability to fool any machine learning model. As the word implies, “foe” refers to a rival, whereas “rival” refers to a foe. In order to strengthen the machine learning models, this section discusses about the weakness of machine learning models and how effectively the misinterpretation occurs during the learning cycle. As definite as it is, existing methods such as creating adversarial models and devising powerful ML computations, frequently ignore semantics and the general skeleton including ML section. This research work develops an adversarial learning calculation by considering the coordinated portrayal by considering all the characteristics and Convolutional Neural Networks (CNN) explicitly. Figuring will most likely express minimal adjustments via data transport represented over positive and negative class markings, as well as a specific subsequent data flow misclassified by CNN. The final results recommend a certain game theory and formative figuring, which obtain incredible favored ensuring about significant learning models against the execution of shortcomings, which are reproduced as attack circumstances against various adversaries.
{"title":"Adversarial Deep Learning Models With Multiple Adversaries","authors":"N. Janapriya, K. Anuradha, V. Srilakshmi","doi":"10.1109/ICIRCA51532.2021.9544889","DOIUrl":"https://doi.org/10.1109/ICIRCA51532.2021.9544889","url":null,"abstract":"Adversarial machine learning calculations handle adversarial instance age, producing bogus data information with the ability to fool any machine learning model. As the word implies, “foe” refers to a rival, whereas “rival” refers to a foe. In order to strengthen the machine learning models, this section discusses about the weakness of machine learning models and how effectively the misinterpretation occurs during the learning cycle. As definite as it is, existing methods such as creating adversarial models and devising powerful ML computations, frequently ignore semantics and the general skeleton including ML section. This research work develops an adversarial learning calculation by considering the coordinated portrayal by considering all the characteristics and Convolutional Neural Networks (CNN) explicitly. Figuring will most likely express minimal adjustments via data transport represented over positive and negative class markings, as well as a specific subsequent data flow misclassified by CNN. The final results recommend a certain game theory and formative figuring, which obtain incredible favored ensuring about significant learning models against the execution of shortcomings, which are reproduced as attack circumstances against various adversaries.","PeriodicalId":245244,"journal":{"name":"2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA)","volume":"458 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116550581","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：