Pub Date : 2021-09-21DOI: 10.1109/aict52120.2021.9628990
B. Rusyn, Y. Obukh, R. Kosarevych, O. Lutsyk, V. Korniy
This paper proposes information technology for analyzing the condition of forest plantations and monitoring the ecological condition of forests. Information technology is based on the proposed approach to automatic localization and recognition of affected trees, and is of great practical importance for environmental monitoring and forestry. A deep learning model has been developed for localization and recognition. This model consists of a detector and separate classifier modules. In order to train and validate the proposed network based on remote sensing images, a training database containing 9000 images was created. Comparison of the proposed model with existing methods is based on characteristics such as accuracy and speed. The accuracy and speed of the proposed recognition system was assessed on a validation sample of images, the size of which is 2000 images. To ensure real-time operation, the software is optimized to work with the GPU. The results of the work and the developed software are used in remote monitoring and classification systems for environmental monitoring and in applied tasks of forestry.
{"title":"Information System for Analysis of Forest Plantations and Monitoring of Ecological Condition","authors":"B. Rusyn, Y. Obukh, R. Kosarevych, O. Lutsyk, V. Korniy","doi":"10.1109/aict52120.2021.9628990","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628990","url":null,"abstract":"This paper proposes information technology for analyzing the condition of forest plantations and monitoring the ecological condition of forests. Information technology is based on the proposed approach to automatic localization and recognition of affected trees, and is of great practical importance for environmental monitoring and forestry. A deep learning model has been developed for localization and recognition. This model consists of a detector and separate classifier modules. In order to train and validate the proposed network based on remote sensing images, a training database containing 9000 images was created. Comparison of the proposed model with existing methods is based on characteristics such as accuracy and speed. The accuracy and speed of the proposed recognition system was assessed on a validation sample of images, the size of which is 2000 images. To ensure real-time operation, the software is optimized to work with the GPU. The results of the work and the developed software are used in remote monitoring and classification systems for environmental monitoring and in applied tasks of forestry.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"196 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115149874","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628985
O. Lavrynenko, R. Odarchenko, G. Konakhovych, A. Taranenko, D. Bakhtiiarov, T. Dyka
The result of this work is to solve the current scientific and practical problem of developing and researching new effective methods of semantic coding of speech signals. A known method of semantic coding of speech signals based on mel-frequency cepstral coefficients, which does not comply with the condition of adaptability to the studied signal, which is a significant disadvantage of the existing method. A method of semantic coding of speech signals based on empirical wavelet transform is developed, which constructs sets of adaptive Meyer wavelet filters with subsequent application of Hilbert spectral analysis to find instantaneous amplitudes and frequencies of functions of internal empirical modes, which will determine semantic efficiency coding. It is proposed to use the method of adaptive empirical wavelet transform in problems of multiplescale analysis and semantic coding of speech signals, which will increase the efficiency of spectral analysis by decomposing high- frequency speech oscillations into its low-frequency components, namely internal empirical modes. An algorithm for semantic coding of speech signals based on empirical wavelet transform and its software implementation in the MATLAB R2020b programming language has been developed.
{"title":"Method of Semantic Coding of Speech Signals based on Empirical Wavelet Transform","authors":"O. Lavrynenko, R. Odarchenko, G. Konakhovych, A. Taranenko, D. Bakhtiiarov, T. Dyka","doi":"10.1109/aict52120.2021.9628985","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628985","url":null,"abstract":"The result of this work is to solve the current scientific and practical problem of developing and researching new effective methods of semantic coding of speech signals. A known method of semantic coding of speech signals based on mel-frequency cepstral coefficients, which does not comply with the condition of adaptability to the studied signal, which is a significant disadvantage of the existing method. A method of semantic coding of speech signals based on empirical wavelet transform is developed, which constructs sets of adaptive Meyer wavelet filters with subsequent application of Hilbert spectral analysis to find instantaneous amplitudes and frequencies of functions of internal empirical modes, which will determine semantic efficiency coding. It is proposed to use the method of adaptive empirical wavelet transform in problems of multiplescale analysis and semantic coding of speech signals, which will increase the efficiency of spectral analysis by decomposing high- frequency speech oscillations into its low-frequency components, namely internal empirical modes. An algorithm for semantic coding of speech signals based on empirical wavelet transform and its software implementation in the MATLAB R2020b programming language has been developed.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"358 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114127643","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper will provide a summary of the drivers which makes changes within the media industry and film production based on the progress of Information and Communications Technologies. As for now - the entertainment industry is constantly growing, based on reaching a more international audience and using new, creative ways of developing and delivering content. Despite the fact that entertainment is more close to Art - Information and Communications Technologies makes big and valuable changes to the industry, not only in terms of new devices, hardware, and software but also in the ways of creatively using those resources. Some of the examples in the article shows how technologies can change completely the ways how some instruments were used before, and how more sufficient and handier they can be only because new technologies opened new opportunities. Merging of the industries like game, movie and informational technologies is a matter of time and IT creations like “Unreal Engine” blurs the lines between gaming and filming, taking computer graphics and 3D modeling to a new level. Other big changes to the Art can bring Artificial Intelligence, Machine Learning, and Big Data instruments, although it’s still not so trusted by big production companies and film studios. But the more Information and Communications Technologies are evolving, the more Media Industry is changing.
{"title":"Informational Technologies in Film Production - How ICT shaping Media Industry","authors":"Aidiye Aidarbekov, Gulmira Shakhmetova, Kamshat Asmaganbetova, Zamira Bekish, Abzal E. Kyzyrkanov, Abylay Salimzhanov","doi":"10.1109/aict52120.2021.9628901","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628901","url":null,"abstract":"This paper will provide a summary of the drivers which makes changes within the media industry and film production based on the progress of Information and Communications Technologies. As for now - the entertainment industry is constantly growing, based on reaching a more international audience and using new, creative ways of developing and delivering content. Despite the fact that entertainment is more close to Art - Information and Communications Technologies makes big and valuable changes to the industry, not only in terms of new devices, hardware, and software but also in the ways of creatively using those resources. Some of the examples in the article shows how technologies can change completely the ways how some instruments were used before, and how more sufficient and handier they can be only because new technologies opened new opportunities. Merging of the industries like game, movie and informational technologies is a matter of time and IT creations like “Unreal Engine” blurs the lines between gaming and filming, taking computer graphics and 3D modeling to a new level. Other big changes to the Art can bring Artificial Intelligence, Machine Learning, and Big Data instruments, although it’s still not so trusted by big production companies and film studios. But the more Information and Communications Technologies are evolving, the more Media Industry is changing.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114507911","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628948
M. Iavich, G. Iashvili, A. Gagnidze, R. Odarchenko
Content filtering tools are fairly well established today. They allow us to find content that is relevant to our needs and interests. Many web platforms such as travel databases, online shops, educational systems and others use recommender tools to filter the content. For these systems, the issue of user security is quite acute on the agenda as hardware based systems are very often attacked by hackers. Many efficient results can be achieved by incorporating machine learning mechanisms into the recommendation systems. This will notably increase hardware-based systems security level. Information retrieval systems that rely on content-based mechanisms provide users with the relevant information and at the same time analyze their behavior. This is done to Figure out how effective the search was for the user. A model of a content-based recommendation system can be incorporated into existing security mechanisms. The network system developed in our study uses the integrated filtering mechanism to identify the frequency of the term received from the user input and it delivers the relevant content to the end user. This system has the advantage of the interacting with specific user cases. It should be noted that the content-based systems may not be fully functional by taking more complex interactions and user actions. This fact is the important field of the research in the world.
{"title":"Use of Content-Filtering Method for Hardware Vulnerabilities Identification System","authors":"M. Iavich, G. Iashvili, A. Gagnidze, R. Odarchenko","doi":"10.1109/aict52120.2021.9628948","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628948","url":null,"abstract":"Content filtering tools are fairly well established today. They allow us to find content that is relevant to our needs and interests. Many web platforms such as travel databases, online shops, educational systems and others use recommender tools to filter the content. For these systems, the issue of user security is quite acute on the agenda as hardware based systems are very often attacked by hackers. Many efficient results can be achieved by incorporating machine learning mechanisms into the recommendation systems. This will notably increase hardware-based systems security level. Information retrieval systems that rely on content-based mechanisms provide users with the relevant information and at the same time analyze their behavior. This is done to Figure out how effective the search was for the user. A model of a content-based recommendation system can be incorporated into existing security mechanisms. The network system developed in our study uses the integrated filtering mechanism to identify the frequency of the term received from the user input and it delivers the relevant content to the end user. This system has the advantage of the interacting with specific user cases. It should be noted that the content-based systems may not be fully functional by taking more complex interactions and user actions. This fact is the important field of the research in the world.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126062005","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628932
A. Kuznetsov, Serhii Datsenko, Yuriy Gorbenko, Tetiana Chupilko, M. Korneyev, Victoria Klym
User authentication technologies are an important component of modern cybersecurity. These technologies are designed to verify the correctness and validity, for example, by verifying the authenticity of entered passwords, identifiers, certain biometric features, and so on. We consider biometric authentication techniques for the human face. Two techniques are used to process biometric features in our work: convolutional neural networks and histograms of directed gradients. We conduct experimental research to study these techniques and compare the results of both in terms of authentication accuracy and the speed of processing biometric images. The obtained results will be useful in the development and implementation of cybersecurity applications that use biometric authentication
{"title":"Experimental Researches of Biometric Authentication Using Convolutional Neural Networks and Histograms of Oriented Graphs","authors":"A. Kuznetsov, Serhii Datsenko, Yuriy Gorbenko, Tetiana Chupilko, M. Korneyev, Victoria Klym","doi":"10.1109/aict52120.2021.9628932","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628932","url":null,"abstract":"User authentication technologies are an important component of modern cybersecurity. These technologies are designed to verify the correctness and validity, for example, by verifying the authenticity of entered passwords, identifiers, certain biometric features, and so on. We consider biometric authentication techniques for the human face. Two techniques are used to process biometric features in our work: convolutional neural networks and histograms of directed gradients. We conduct experimental research to study these techniques and compare the results of both in terms of authentication accuracy and the speed of processing biometric images. The obtained results will be useful in the development and implementation of cybersecurity applications that use biometric authentication","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125250664","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628973
V. Barannik, D. Barannik, Yevhenii Sidchenko, O. Ignatyev, O. Slobodyanyuk, Lazuta Roman
This article contains information on the analysis of existing steganographic methods of hiding information in order to create a new one. Media files, including video ones, are considered as a container for embedding information in this research. The aim of the research is to develop new steganographic method of data embedding. Due to the fact of using outlines as a container, the developed method allows embedding a large amount of information without significantly changing the appearance of the container. The presence of several levels of abstraction from specific values increases the security of built-in data from external interference.
{"title":"Method of Masking Information in the Contours of Video Images","authors":"V. Barannik, D. Barannik, Yevhenii Sidchenko, O. Ignatyev, O. Slobodyanyuk, Lazuta Roman","doi":"10.1109/aict52120.2021.9628973","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628973","url":null,"abstract":"This article contains information on the analysis of existing steganographic methods of hiding information in order to create a new one. Media files, including video ones, are considered as a container for embedding information in this research. The aim of the research is to develop new steganographic method of data embedding. Due to the fact of using outlines as a container, the developed method allows embedding a large amount of information without significantly changing the appearance of the container. The presence of several levels of abstraction from specific values increases the security of built-in data from external interference.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132556951","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628923
Saruhan Furkan Senturk, Muhammet Fatih Donmez, Huseyin Kaan Gulmez, Mehmet Gul, Pınar Kırcı
Environmental pollution is an important factor both for the economy and for a livable environment. For this reason, a lot of work has been done recently to reduce environmental pollution, to increase recycling, to protect the environment and to prevent economic wear. Many municipalities around the world put recycling bins in certain points of cities, districts and even neighborhoods. however, these boxes are designed only to dispose of materials that can be recycled by human hand, according to certain containers. Whichever section is reserved for which waste, people separate their garbage and throw their garbage in that section. Despite this, it can be seen that different wastes are in different reservoirs and a manpower is needed to separate them again. The project will separate recyclable materials according to their type (glass, paper, nylon, metal) into different containers using sensors and smart systems.
{"title":"A Low-Cost Intelligent Metal Recycling Machine","authors":"Saruhan Furkan Senturk, Muhammet Fatih Donmez, Huseyin Kaan Gulmez, Mehmet Gul, Pınar Kırcı","doi":"10.1109/aict52120.2021.9628923","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628923","url":null,"abstract":"Environmental pollution is an important factor both for the economy and for a livable environment. For this reason, a lot of work has been done recently to reduce environmental pollution, to increase recycling, to protect the environment and to prevent economic wear. Many municipalities around the world put recycling bins in certain points of cities, districts and even neighborhoods. however, these boxes are designed only to dispose of materials that can be recycled by human hand, according to certain containers. Whichever section is reserved for which waste, people separate their garbage and throw their garbage in that section. Despite this, it can be seen that different wastes are in different reservoirs and a manpower is needed to separate them again. The project will separate recyclable materials according to their type (glass, paper, nylon, metal) into different containers using sensors and smart systems.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"265 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127707386","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628981
Valerii Lovkin, A. Oliinyk, Tetiana Fedoronchak, Yurii Lukashenko
Medical diagnosis system needs prediction data on concentration of air pollutants to support making of personal decisions about outdoor activities for a day. It should extend decision made by doctor concerning medical diagnosis to decisions made by patient. Information model of outdoor air pollution prediction was presented. Information model is based on prediction model which was created and trained for prediction of nitrogen dioxide concentration. Prediction model was created using recurrent neural network based on long short term memory architecture. Experimental investigation was performed using dataset collected in Madrid during period from 2001 to 2020. Experimental investigation approved efficiency of the developed model. The created information model was developed as separate module of nitrogen dioxide concentration prediction inside medical diagnosis system.
{"title":"Information Model of Outdoor Air Pollution Prediction for Medical Diagnosis System","authors":"Valerii Lovkin, A. Oliinyk, Tetiana Fedoronchak, Yurii Lukashenko","doi":"10.1109/aict52120.2021.9628981","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628981","url":null,"abstract":"Medical diagnosis system needs prediction data on concentration of air pollutants to support making of personal decisions about outdoor activities for a day. It should extend decision made by doctor concerning medical diagnosis to decisions made by patient. Information model of outdoor air pollution prediction was presented. Information model is based on prediction model which was created and trained for prediction of nitrogen dioxide concentration. Prediction model was created using recurrent neural network based on long short term memory architecture. Experimental investigation was performed using dataset collected in Madrid during period from 2001 to 2020. Experimental investigation approved efficiency of the developed model. The created information model was developed as separate module of nitrogen dioxide concentration prediction inside medical diagnosis system.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116902498","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628939
E. Lavrov, O. Siryk, Y. Chybiriak, L. Danilova, V. Nahornyi, Svetlana Vakal
Problems of ensuring the reliable and efficient functioning of dialogue automated systems are considered. The necessity of special attention to the so-called “human factor” is shown. A meaningful analysis of the mechanisms of adaptation of the dialogue system to the human operator is carried out. An approach to the construction of a dialogue model as a process of changing the states of the “human-computer” system is proposed. It is proposed to formalize the problem of adaptive dialogue control in the form of an optimization problem using semi-Markov decision-making processes. The advantage of the proposed approach is the reduction of the problem to a linear programming problem, which is easily implemented in the construction of real dialogue control systems. The approbation in the systems of e-business, technological process management, e-learning has proved the constructiveness of the approach.
{"title":"A Model for the Organization of Adaptive Dialogue Interaction “Man-Computer” Taking into Account the Requirements of Reliability and Efficiency","authors":"E. Lavrov, O. Siryk, Y. Chybiriak, L. Danilova, V. Nahornyi, Svetlana Vakal","doi":"10.1109/aict52120.2021.9628939","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628939","url":null,"abstract":"Problems of ensuring the reliable and efficient functioning of dialogue automated systems are considered. The necessity of special attention to the so-called “human factor” is shown. A meaningful analysis of the mechanisms of adaptation of the dialogue system to the human operator is carried out. An approach to the construction of a dialogue model as a process of changing the states of the “human-computer” system is proposed. It is proposed to formalize the problem of adaptive dialogue control in the form of an optimization problem using semi-Markov decision-making processes. The advantage of the proposed approach is the reduction of the problem to a linear programming problem, which is easily implemented in the construction of real dialogue control systems. The approbation in the systems of e-business, technological process management, e-learning has proved the constructiveness of the approach.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117181516","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 : 2021-09-21DOI: 10.1109/aict52120.2021.9628974
Serhii Leoshcheko, A. Oliinyk, S. Subbotin, Mykyta Zaiko
During performing complex synthesis (structural and parametric) of artificial neural networks (ANN) of complex topologies (for example, recurrent neural networks: RNN and deep neural networks: DNN), neuroevolution is a promising approach. After all, methods of neuroevolution synthesis allow to obtain the appropriate topology and parameters of the neuromodel with less involvement of an external expert, by gradually increasing, increasing the complexity and changing the network parameters. However, neuroevolution synthesis has a significant execution time. There are also sometimes problems with local extremes. That’s why the urgent task is to develop modifications that can be introduced point-by-point during the synthesis process in order to overcome disadvantages that were listed and improve results of using such methods.
{"title":"Mechanisms of Fine Tuning of Neuroevolutionary Synthesis of Artificial Neural Networks","authors":"Serhii Leoshcheko, A. Oliinyk, S. Subbotin, Mykyta Zaiko","doi":"10.1109/aict52120.2021.9628974","DOIUrl":"https://doi.org/10.1109/aict52120.2021.9628974","url":null,"abstract":"During performing complex synthesis (structural and parametric) of artificial neural networks (ANN) of complex topologies (for example, recurrent neural networks: RNN and deep neural networks: DNN), neuroevolution is a promising approach. After all, methods of neuroevolution synthesis allow to obtain the appropriate topology and parameters of the neuromodel with less involvement of an external expert, by gradually increasing, increasing the complexity and changing the network parameters. However, neuroevolution synthesis has a significant execution time. There are also sometimes problems with local extremes. That’s why the urgent task is to develop modifications that can be introduced point-by-point during the synthesis process in order to overcome disadvantages that were listed and improve results of using such methods.","PeriodicalId":375013,"journal":{"name":"2021 IEEE 4th International Conference on Advanced Information and Communication Technologies (AICT)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114329769","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：