Monitoring of heart rate (HR) and its dynamics is necessary in ambulatory and telemedicine for diagnosis and treatment of diseases. Remote photoplethysmography (rPPG) allows for HR and its dynamics monitoring by video analysis of facial skin blood flow, and is ofparticular importance for patients with delicate skin such as infants, the elderly, or those with severe burn injuries. Unlike other contactless HR measurement methods, rPPG does not require special hardware, but allows to monitor HR on the basis of analyzing a sequence of video images of a person's face. rPPG involves a multi-step process including signal extraction, dimensionality reduction to estimate the photoplethysmographic (PPG) signal, and subsequent HR estimation. However, the presence of high-amplitude spikes due to subject movement, facial expressions, lighting fluctuations, video compression artifacts, ROI tracking errors, among others, can interfere the useful PPG signal, leading to inaccurate HR estimations. A method has been developed that allows to increase the accuracy of HR estimation in the rPPG problem due to its robustness to interferences inevitable during video recording. The proposed approach tackles the rPPG challenge by applying a pre-processing approximation of the signal finite difference using a single-layer neural network with radial basis function (RBF) inner layer. Transitioning to the signal finite difference helps reduce the amplitudes of irrelevant low-frequency peaks within the HR search range, thus avoiding their masking effect on the HR-related spectral peaks. The neural network's RBF approximation further diminishes irrelevant high-frequency spectral peaks when the number of RBF nodes is less than half the signal sample count. The correctness of the solutions is confirmed by numerical experiments carried out on the Mahnob-HCI public database.
{"title":"Method of Remote Photoplethysmography Robust to Interference in Video Registration of Human Facial Skin","authors":"M. V. Kopeliovich, I. V. Shcherban","doi":"10.17587/it.30.357-366","DOIUrl":"https://doi.org/10.17587/it.30.357-366","url":null,"abstract":"Monitoring of heart rate (HR) and its dynamics is necessary in ambulatory and telemedicine for diagnosis and treatment of diseases. Remote photoplethysmography (rPPG) allows for HR and its dynamics monitoring by video analysis of facial skin blood flow, and is ofparticular importance for patients with delicate skin such as infants, the elderly, or those with severe burn injuries. Unlike other contactless HR measurement methods, rPPG does not require special hardware, but allows to monitor HR on the basis of analyzing a sequence of video images of a person's face. rPPG involves a multi-step process including signal extraction, dimensionality reduction to estimate the photoplethysmographic (PPG) signal, and subsequent HR estimation. However, the presence of high-amplitude spikes due to subject movement, facial expressions, lighting fluctuations, video compression artifacts, ROI tracking errors, among others, can interfere the useful PPG signal, leading to inaccurate HR estimations. A method has been developed that allows to increase the accuracy of HR estimation in the rPPG problem due to its robustness to interferences inevitable during video recording. The proposed approach tackles the rPPG challenge by applying a pre-processing approximation of the signal finite difference using a single-layer neural network with radial basis function (RBF) inner layer. Transitioning to the signal finite difference helps reduce the amplitudes of irrelevant low-frequency peaks within the HR search range, thus avoiding their masking effect on the HR-related spectral peaks. The neural network's RBF approximation further diminishes irrelevant high-frequency spectral peaks when the number of RBF nodes is less than half the signal sample count. The correctness of the solutions is confirmed by numerical experiments carried out on the Mahnob-HCI public database.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141654218","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The paper presents the experiments' results on the study variants of neural network architectures for predicting wind speed based on meteorological time series for Northern Dagestan. When working with data and models, modern software tools of the Python programming language for Data Science are used, such as Keras — a library for modeling neural networks, Pandas — a library for processing tabular data, AutoKeras — a library for automatically generating a neural network by dataset, TabGan — a library for expansion a dataset with artificial data. As initial data, regular observations at the Kochubey (Northern Dagestan) meteorological station for the period 20112022 were taken with a frequency of generalization of measurements 8 times a day. The original semi-structured data is pre-processed and reduced to a structured CSV dataset format. The task of predicting wind speed is reduced to the task of classification, in which it is not the wind speed itself that is predicted, but the class number in accordance with the gradation. From the point of view of considering wind speed as a renewable energy resource, three classes with gradations are accepted: class 0: 0—3 m/s (quiet), class 1: 4—7 m/s (average wind sufficient for optimal wind turbine operating), class 2: 8 m/s and higher (strong wind). When performing experiments, the influence value on the prediction accuracy from several aspects were analyzed, such as: the time data block length, the neural network architecture, the transformation tabular features to normal form or to categorical form, expansion dataset by artificial data, the layout of the verification and test samples, the imbalance of classes, various meteorological parameters as features.
{"title":"Investigation Neural Network Models for Wind Speed Prediction Based on Meteorological Observations in Northern Dagestan","authors":"D. N. Kobzarenko, A. M. Kamilova","doi":"10.17587/it.30.350-356","DOIUrl":"https://doi.org/10.17587/it.30.350-356","url":null,"abstract":"The paper presents the experiments' results on the study variants of neural network architectures for predicting wind speed based on meteorological time series for Northern Dagestan. When working with data and models, modern software tools of the Python programming language for Data Science are used, such as Keras — a library for modeling neural networks, Pandas — a library for processing tabular data, AutoKeras — a library for automatically generating a neural network by dataset, TabGan — a library for expansion a dataset with artificial data. As initial data, regular observations at the Kochubey (Northern Dagestan) meteorological station for the period 20112022 were taken with a frequency of generalization of measurements 8 times a day. The original semi-structured data is pre-processed and reduced to a structured CSV dataset format. The task of predicting wind speed is reduced to the task of classification, in which it is not the wind speed itself that is predicted, but the class number in accordance with the gradation. From the point of view of considering wind speed as a renewable energy resource, three classes with gradations are accepted: class 0: 0—3 m/s (quiet), class 1: 4—7 m/s (average wind sufficient for optimal wind turbine operating), class 2: 8 m/s and higher (strong wind). When performing experiments, the influence value on the prediction accuracy from several aspects were analyzed, such as: the time data block length, the neural network architecture, the transformation tabular features to normal form or to categorical form, expansion dataset by artificial data, the layout of the verification and test samples, the imbalance of classes, various meteorological parameters as features.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141653042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The study is devoted to improving evacuation strategies in emergency situations based on the application of the evolutionary algorithm Mole-rat (mole colony algorithm) (MRA). The work is based on a modified dynamic obstacle avoidance mechanism designed to improve the efficiency of the basic MRA algorithm in the context of evacuation from hazardous areas. In the context of increasing emergencies and crises, providing effective evacuation strategies becomes a pressing and important task. The implementation of the MRA algorithm with a modified dynamic obstacle avoidance mechanism is a relevant approach that will improve the efficiency and safety of real-time evacuation.
{"title":"Modified Evolutionary Algorithm Mole-Rat with an Adaptive Mechanism for Dynamic Obstacle Avoidance in Emergency Situations","authors":"V. V. Kureichik, V. Danilchenko","doi":"10.17587/it.30.342-349","DOIUrl":"https://doi.org/10.17587/it.30.342-349","url":null,"abstract":"The study is devoted to improving evacuation strategies in emergency situations based on the application of the evolutionary algorithm Mole-rat (mole colony algorithm) (MRA). The work is based on a modified dynamic obstacle avoidance mechanism designed to improve the efficiency of the basic MRA algorithm in the context of evacuation from hazardous areas. In the context of increasing emergencies and crises, providing effective evacuation strategies becomes a pressing and important task. The implementation of the MRA algorithm with a modified dynamic obstacle avoidance mechanism is a relevant approach that will improve the efficiency and safety of real-time evacuation.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141652258","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The study proposes an approach to diagnostics of diseases based on the analytic hierarchy process, which is as follows: by the method of pairwise comparisons, the weight structures of the symptoms of the alleged disease and the patient's disease are investigated, then the measure of the differences between these structures is evaluated, based on this assessment, we can talk about the reliability of the patient's diagnosis. Unlike the classical analytic hierarchy process, there are no alternatives in the developed modification, and diagnostics is carried out only by comparing the weight structures of the symptoms of the alleged disease and the patient's disease, that is, the approach does not imply the establishment of one of the alternative diagnoses, but only confirms (or refutes) the presence of the patient's disease. The method is tested on the basis of a clinical case of a mental disorder, the index of structural differences showed the identity of the weight structures of the symptoms of the alleged disease and the patient's disease, which confirmed the reliability of the patient's diagnosis. The method can be used to assess the reliability of the diagnosis, and, if repeated, the clinical dynamics of the disease in the course of therapy.
{"title":"Analytic Hierarchy Process in Diagnostics of Diseases","authors":"S. A. Tarasova, V. V. Tarasov","doi":"10.17587/it.30.367-371","DOIUrl":"https://doi.org/10.17587/it.30.367-371","url":null,"abstract":"The study proposes an approach to diagnostics of diseases based on the analytic hierarchy process, which is as follows: by the method of pairwise comparisons, the weight structures of the symptoms of the alleged disease and the patient's disease are investigated, then the measure of the differences between these structures is evaluated, based on this assessment, we can talk about the reliability of the patient's diagnosis. Unlike the classical analytic hierarchy process, there are no alternatives in the developed modification, and diagnostics is carried out only by comparing the weight structures of the symptoms of the alleged disease and the patient's disease, that is, the approach does not imply the establishment of one of the alternative diagnoses, but only confirms (or refutes) the presence of the patient's disease. The method is tested on the basis of a clinical case of a mental disorder, the index of structural differences showed the identity of the weight structures of the symptoms of the alleged disease and the patient's disease, which confirmed the reliability of the patient's diagnosis. The method can be used to assess the reliability of the diagnosis, and, if repeated, the clinical dynamics of the disease in the course of therapy.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141652166","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The paper identifies three main approaches to obtaining Hadamard matrices: search using combinatorial methods, calculation with control based on the theory of dynamical systems and the construction of matrices of fixed structures. For the search and construction of Hadamard matrices, the main tool of the source material is the generation of random sequences. The issues of fixing the structures of Hadamard matrices in the form of a bicyclic construction are considered. To obtain such high-order matrices, important procedures are the generation, filtering and selection of such pairs of sequences that an orthogonal matrix of order n could be constructed from the cyclic matrices of order n/2 obtained on their basis. There is a significant influence of the quality of generated random sequences on the construction time of bicyclic matrices. The results of the first experiments with 1 million random sequences of length 100 generated on a quantum generator based on the interference effect of laser pulses with a random phase are presented. In particular, previously unknown Hadamard matrices of orders up to 100 bicyclic structures and maximum determinant matrices on non-Hadamard orders were obtained in a computer experiment.
{"title":"On Quantum Generation of Random Sequences as the Basis for Constructing Bicyclic Orthogonal Matrices","authors":"A. M. Sergeev","doi":"10.17587/it.30.331-335","DOIUrl":"https://doi.org/10.17587/it.30.331-335","url":null,"abstract":"The paper identifies three main approaches to obtaining Hadamard matrices: search using combinatorial methods, calculation with control based on the theory of dynamical systems and the construction of matrices of fixed structures. For the search and construction of Hadamard matrices, the main tool of the source material is the generation of random sequences. The issues of fixing the structures of Hadamard matrices in the form of a bicyclic construction are considered. To obtain such high-order matrices, important procedures are the generation, filtering and selection of such pairs of sequences that an orthogonal matrix of order n could be constructed from the cyclic matrices of order n/2 obtained on their basis. There is a significant influence of the quality of generated random sequences on the construction time of bicyclic matrices. The results of the first experiments with 1 million random sequences of length 100 generated on a quantum generator based on the interference effect of laser pulses with a random phase are presented. In particular, previously unknown Hadamard matrices of orders up to 100 bicyclic structures and maximum determinant matrices on non-Hadamard orders were obtained in a computer experiment.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141654359","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The task of assessing student perception of the gaming chatbot with Wordle like puzzles for learning programming is considered. An experiment was conducted, the processing of the results of which made it possible to build 3 regression models describing the influence of factors on the interest, difficulty, and perceived usefulness of the gaming chatbot. Recommendations for using the gaming chatbot are formulated.
{"title":"Models of Interest, Difficulty, and Perceived Usefulness of the Gaming Chatbot with Wordle Like Puzzles for Learning Programming","authors":"A. N. Varnavsky","doi":"10.17587/it.30.372-381","DOIUrl":"https://doi.org/10.17587/it.30.372-381","url":null,"abstract":"The task of assessing student perception of the gaming chatbot with Wordle like puzzles for learning programming is considered. An experiment was conducted, the processing of the results of which made it possible to build 3 regression models describing the influence of factors on the interest, difficulty, and perceived usefulness of the gaming chatbot. Recommendations for using the gaming chatbot are formulated.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141652330","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}
V. I. Lutin, F. A. Desyatirikov, A. S. Belyaev, E. V. Lutina, E. N. Desyatirikova
Under conditions of an unknown shift of the reference points of a continuous signal and its given fragment, an approach to the synthesis of a fragment search algorithm is proposed. The approach is based on the classical apparatus for testing hypotheses of mathematical statistics. A distinctive feature of the approach is taking into account the noise of the signal and fragment reading, and on this basis, the formalization of the problem of choosing the number of channels for monitoring the signal. Thus, in this work, using a probabilistic-statistical method, we synthesized an algorithm for a two-stage procedure for searching for a given segment in a long data sequence, which consists in determining the most probable position of the desired segment and checking whether the found segment matches the given one. Analytical relations are obtained to assess the quality of detection, which take into account the autocorrelation properties of sequences, and detection characteristics are calculated, which determine the sample size of a given segment and a sufficient number of detector channels to ensure the required quality of detection under given observation conditions. A block diagram of a device for making decisions about the most probable hypothesis and determining the average value and standard deviation of the start time of the desired segment is proposed, which allows taking into account the least favorable conditions for the beginning of the signal countdown, namely, exactly in the middle of the specified fragment.
{"title":"Synthesis of an Algorithm for Searching a Given Segment in a Data Sequence of Finite Length","authors":"V. I. Lutin, F. A. Desyatirikov, A. S. Belyaev, E. V. Lutina, E. N. Desyatirikova","doi":"10.17587/it.30.335-341","DOIUrl":"https://doi.org/10.17587/it.30.335-341","url":null,"abstract":"Under conditions of an unknown shift of the reference points of a continuous signal and its given fragment, an approach to the synthesis of a fragment search algorithm is proposed. The approach is based on the classical apparatus for testing hypotheses of mathematical statistics. A distinctive feature of the approach is taking into account the noise of the signal and fragment reading, and on this basis, the formalization of the problem of choosing the number of channels for monitoring the signal. Thus, in this work, using a probabilistic-statistical method, we synthesized an algorithm for a two-stage procedure for searching for a given segment in a long data sequence, which consists in determining the most probable position of the desired segment and checking whether the found segment matches the given one. Analytical relations are obtained to assess the quality of detection, which take into account the autocorrelation properties of sequences, and detection characteristics are calculated, which determine the sample size of a given segment and a sufficient number of detector channels to ensure the required quality of detection under given observation conditions. A block diagram of a device for making decisions about the most probable hypothesis and determining the average value and standard deviation of the start time of the desired segment is proposed, which allows taking into account the least favorable conditions for the beginning of the signal countdown, namely, exactly in the middle of the specified fragment.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-07-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141655239","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The purpose of this work is to review the existing technologies for automated knowledge extraction from scientific publications. The main tasks include an analysis of existing methods for automated knowledge extraction, as well as an overview of various software tools used to solve this problem. The article presents a description of the main approaches to automated knowledge extraction, such as machine learning, natural language processing and the development of various methodologies for building knowledge graphs. An analysis of existing sources showed that the main problems associated with automated knowledge extraction are the need to create a large amount of labeled data, the processing of complex structured data, and the need to develop new algorithms for working with such data.
{"title":"Comparative Review of Tasks, Approaches and Tools for Automated Knowledge Extraction from the Texts of Scientific Publications","authors":"S. N. Ushakov, A. O. Saveliev","doi":"10.17587/it.30.291-299","DOIUrl":"https://doi.org/10.17587/it.30.291-299","url":null,"abstract":"The purpose of this work is to review the existing technologies for automated knowledge extraction from scientific publications. The main tasks include an analysis of existing methods for automated knowledge extraction, as well as an overview of various software tools used to solve this problem. The article presents a description of the main approaches to automated knowledge extraction, such as machine learning, natural language processing and the development of various methodologies for building knowledge graphs. An analysis of existing sources showed that the main problems associated with automated knowledge extraction are the need to create a large amount of labeled data, the processing of complex structured data, and the need to develop new algorithms for working with such data.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141365243","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The article considers the algorithm of search results ranking according to the compliance with user's expectations due to the feedback on the data of search results pre-release, which allowed to reduce the time of search for the necessary information. A numerical experiment aimed at evaluating the effectiveness of the proposed approach on the example of real estate objects is presented. Aggregated data of real estate listings were used as data, and the results of user surveys were used to evaluate the relevance. TOPSIS and PROMETHEE methods were used as pre-ranking algorithms. The ranking results were combined considering their order in the output of both methods. To implement the next step of the algorithm, the pre-release list was partitioned into classes for which users selected a few objects they liked. Machine learning models were trained on the partitioned data. Based on their average accuracy and variance estimates, a naive Bayesian classifier model was selected and used for subsequent computations. The results of further experiments showed the possibility of taking into account personal preferences when organizing search and selection of objects of interest in the absence of personalized statistics on the example of real estate objects. As a result of the experiments, the time of searching a group of results of interest was reduced by 74 % on average.
{"title":"Ranking Search Results Based on User Preferences in the Absence of Personalized Statistics","authors":"A. S. Svitek, L. A. Mylnikov","doi":"10.17587/it.30.307-317","DOIUrl":"https://doi.org/10.17587/it.30.307-317","url":null,"abstract":"The article considers the algorithm of search results ranking according to the compliance with user's expectations due to the feedback on the data of search results pre-release, which allowed to reduce the time of search for the necessary information. A numerical experiment aimed at evaluating the effectiveness of the proposed approach on the example of real estate objects is presented. Aggregated data of real estate listings were used as data, and the results of user surveys were used to evaluate the relevance. TOPSIS and PROMETHEE methods were used as pre-ranking algorithms. The ranking results were combined considering their order in the output of both methods. To implement the next step of the algorithm, the pre-release list was partitioned into classes for which users selected a few objects they liked. Machine learning models were trained on the partitioned data. Based on their average accuracy and variance estimates, a naive Bayesian classifier model was selected and used for subsequent computations. The results of further experiments showed the possibility of taking into account personal preferences when organizing search and selection of objects of interest in the absence of personalized statistics on the example of real estate objects. As a result of the experiments, the time of searching a group of results of interest was reduced by 74 % on average.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141362390","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The issue of increasing the efficiency of software for computing architectures that support vector extensions of the command system is considered. Modern compilers can perform automatic vectorization of calculations, convert programs from a scalar representation to a vector implementation. The paper analyzes the effectiveness of automatic vectorization performed by modern compilers, discusses the problems inherent in automatic vectorization. A new algorithm for vectorization of calculations is proposed, which allows to significantly increasing the efficiency of the resulting software.
{"title":"The New Method for Increasing the Efficiency of Vectorization of BLAS Operations","authors":"V. A. Egunov, A. G. Kravets","doi":"10.17587/it.30.318-328","DOIUrl":"https://doi.org/10.17587/it.30.318-328","url":null,"abstract":"The issue of increasing the efficiency of software for computing architectures that support vector extensions of the command system is considered. Modern compilers can perform automatic vectorization of calculations, convert programs from a scalar representation to a vector implementation. The paper analyzes the effectiveness of automatic vectorization performed by modern compilers, discusses the problems inherent in automatic vectorization. A new algorithm for vectorization of calculations is proposed, which allows to significantly increasing the efficiency of the resulting software.","PeriodicalId":504905,"journal":{"name":"Informacionnye Tehnologii","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2024-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141362884","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：