Sentiment analysis is one of the core fields of Natural Language Processing(NLP). Numerous machine learning and deep learning algorithms have been developed to achieve this task. Generally, deep learning models perform better in this task as they are trained on massive amounts of data. This, however, also poses a disadvantage as collecting sufficient amounts of data is a challenge and training with this data requires devices with high computational power. Word embedding is a vital step in applying machine learning models for NLP tasks. Different word embedding techniques affect the performance of machine learning algorithms. This paper evaluates GloVe, CountVectorizer, and TF-IDF embedding techniques with multiple machine learning models and proves that the right combination of embedding technique and machine learning model(TF-IDF+Logistic Regression: 87.75% accuracy) can achieve nearly the same performance or more as deep learning models (LSTM: 87.89%).
{"title":"Performance Evaluation of Different Word Embedding Techniques Across Machine Learning and Deep Learning Models","authors":"Tanmoy Mazumder, Shawan Das, Md. Hasibur Rahman, Tanjina Helaly, Tanmoy Sarkar Pias","doi":"10.1109/ICCIT57492.2022.10055572","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055572","url":null,"abstract":"Sentiment analysis is one of the core fields of Natural Language Processing(NLP). Numerous machine learning and deep learning algorithms have been developed to achieve this task. Generally, deep learning models perform better in this task as they are trained on massive amounts of data. This, however, also poses a disadvantage as collecting sufficient amounts of data is a challenge and training with this data requires devices with high computational power. Word embedding is a vital step in applying machine learning models for NLP tasks. Different word embedding techniques affect the performance of machine learning algorithms. This paper evaluates GloVe, CountVectorizer, and TF-IDF embedding techniques with multiple machine learning models and proves that the right combination of embedding technique and machine learning model(TF-IDF+Logistic Regression: 87.75% accuracy) can achieve nearly the same performance or more as deep learning models (LSTM: 87.89%).","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"42 11-12","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132496730","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10055683
Wasi Mashrur, Shahriar Bin Salim, Sunjida Sultana, Md. Soyaeb Hasan, Md. Akhter Uz Zaman, K. M. Zahidur Rahman, Md Rafiqul Islam
In this paper, the impact of Extended Back Gate (EBG) length on GaAs based DG-JLMOSFET is simulated to analyze its superior behaviors in contrast with conventional DG- JLMOSFETs. For determining the optimal performance of EBG in DG-JLMOSFET, the back gate is extended symmetrically from gate towards source and drain sides for several distinct lengths ranging from 10 nm to 20 nm. For both top and back gates HfO2 is taken as the gate oxide material and the oxide thickness is considered as 1 nm. For a fixed channel length of 10 nm, the suggested model displays that when gate length is increased the impact of the drain voltage on the drain current is diminished resulting significant decrease in OFF-state current with a larger Ion/Ioff ratio of ~ 109. In fact, this leads to a reduced drain induced barrier lowering. Moreover, numerous simulated results from SILVACO ATLAS TCAD offers larger drain current as well as lower subthreshold swing of 67.5 mV/Dec for the proposed model. Due to its superior performance over traditional DG-JLMOSFET, the proposed structure can be deployed effectively in the near future.
{"title":"Effect of Extended Back Gate in GaAs Based DG- JLMOSFET","authors":"Wasi Mashrur, Shahriar Bin Salim, Sunjida Sultana, Md. Soyaeb Hasan, Md. Akhter Uz Zaman, K. M. Zahidur Rahman, Md Rafiqul Islam","doi":"10.1109/ICCIT57492.2022.10055683","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055683","url":null,"abstract":"In this paper, the impact of Extended Back Gate (EBG) length on GaAs based DG-JLMOSFET is simulated to analyze its superior behaviors in contrast with conventional DG- JLMOSFETs. For determining the optimal performance of EBG in DG-JLMOSFET, the back gate is extended symmetrically from gate towards source and drain sides for several distinct lengths ranging from 10 nm to 20 nm. For both top and back gates HfO2 is taken as the gate oxide material and the oxide thickness is considered as 1 nm. For a fixed channel length of 10 nm, the suggested model displays that when gate length is increased the impact of the drain voltage on the drain current is diminished resulting significant decrease in OFF-state current with a larger Ion/Ioff ratio of ~ 109. In fact, this leads to a reduced drain induced barrier lowering. Moreover, numerous simulated results from SILVACO ATLAS TCAD offers larger drain current as well as lower subthreshold swing of 67.5 mV/Dec for the proposed model. Due to its superior performance over traditional DG-JLMOSFET, the proposed structure can be deployed effectively in the near future.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132518637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10055404
Tithi Paul
A list of components can be arranged in a certain order using a sorting algorithm, which is a fundamental concept in computer science. The temporal complexity of the two fundamental and widely used sorting algorithms, Bubble sort and Insertion sort is $mathcal{O}left( {{N^2}} right)$, where N is the total number of items. When it comes to sorting a specific amount of items, it is superior. However, by adding more parts to its quadratic complexity, it loses efficiency. Because of this, it is less frequently employed in computer science’s practical and real-world applications, despite being widely utilized as a subroutine in other areas. Numerous extension techniques for the insertion sort and bubble sort algorithms have been put out in the literature, but none of them tries to combine the two to create a combination algorithm like ours. The bubble and insertion sort method was modified in this study, and its computational complexity was estimated to be $mathcal{O}(Nsqrt N )$. The technique begins by dividing the input array into a few pieces, sorting each of the blocks using a modified bubble sort, and then merging all of the blocks together using a modified insertion sort. The suggested bubble and insertion sort outperform traditional bubble and insertion sorting as well as all other sorting algorithms with a computational complexity of $mathcal{O}left( {{N^2}} right)$.
一个组件列表可以使用排序算法按照一定的顺序排列，这是计算机科学中的一个基本概念。冒泡排序(Bubble sort)和插入排序(insert sort)这两种基本且广泛使用的排序算法的时间复杂度为$mathcal{O}left( {{N^2}} right)$，其中N为项目总数。当涉及到分类特定数量的物品时，它是优越的。然而，通过增加二次复杂度的部分，它失去了效率。正因为如此，尽管在其他领域作为子例程被广泛使用，但它在计算机科学的实际和实际应用中较少使用。文献中已经提出了许多插入排序和冒泡排序算法的扩展技术，但没有一个试图将两者结合起来创建像我们这样的组合算法。本文对气泡插入排序方法进行了改进，估计其计算复杂度为$mathcal{O}(Nsqrt N )$。该技术首先将输入数组分成几个部分，使用修改后的冒泡排序对每个块进行排序，然后使用修改后的插入排序将所有块合并在一起。建议的气泡和插入排序优于传统的气泡和插入排序以及所有其他排序算法，计算复杂度为$mathcal{O}left( {{N^2}} right)$。
{"title":"Enhancement of Bubble and Insertion Sort Algorithm Using Block Partitioning","authors":"Tithi Paul","doi":"10.1109/ICCIT57492.2022.10055404","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055404","url":null,"abstract":"A list of components can be arranged in a certain order using a sorting algorithm, which is a fundamental concept in computer science. The temporal complexity of the two fundamental and widely used sorting algorithms, Bubble sort and Insertion sort is $mathcal{O}left( {{N^2}} right)$, where N is the total number of items. When it comes to sorting a specific amount of items, it is superior. However, by adding more parts to its quadratic complexity, it loses efficiency. Because of this, it is less frequently employed in computer science’s practical and real-world applications, despite being widely utilized as a subroutine in other areas. Numerous extension techniques for the insertion sort and bubble sort algorithms have been put out in the literature, but none of them tries to combine the two to create a combination algorithm like ours. The bubble and insertion sort method was modified in this study, and its computational complexity was estimated to be $mathcal{O}(Nsqrt N )$. The technique begins by dividing the input array into a few pieces, sorting each of the blocks using a modified bubble sort, and then merging all of the blocks together using a modified insertion sort. The suggested bubble and insertion sort outperform traditional bubble and insertion sorting as well as all other sorting algorithms with a computational complexity of $mathcal{O}left( {{N^2}} right)$.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"82 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131877992","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10055001
Md Momenul Haque, S. Paul, Rakhi Rani Paul, Mirza A. F. M. Rashidul Hasan, Sultan Fahim, S. Islam
South Asia countries like Bangladesh, India, and Pakistan have a large number of fuel filling stations that use centralized payment transaction systems. In some cases, this fuel filling station uses the hand cash payment system which is not secured and time-consuming. Each transaction takes more than five minutes to process. For that reason, in some cases, customers face the huge hassle of standing in a long line and waiting for their turn. Not only that, there are high possibilities of fraud activities and robbery being occur for large amounts of the payment transaction. To solve this problem we propose a blockchain-based payment transaction method for fuel filling stations. Here we use the decentralized open ledger infrastructure and proof-of-work to approve each transaction block. Every transaction between the customer and the filling station authority is completed through a digital wallet which is fully secured, fast, and transparent. Comparing to the bank payment transaction system our proposed method is decentralized and has low transaction fees applied in every transaction. This transaction process is free from third-party involvement and all transactions are immutable. For that reason no issues of customer trust and safe from fraud activities in a large number of payment transactions. Our proposed payment transaction method can play an important part to handle large amounts of transactions and provide transaction security for increasing the number of fuel filling stations in South Asia's most populated country.
{"title":"A Blockchain-Based Secure Payment System for Vehicle Fuel Filling Station","authors":"Md Momenul Haque, S. Paul, Rakhi Rani Paul, Mirza A. F. M. Rashidul Hasan, Sultan Fahim, S. Islam","doi":"10.1109/ICCIT57492.2022.10055001","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055001","url":null,"abstract":"South Asia countries like Bangladesh, India, and Pakistan have a large number of fuel filling stations that use centralized payment transaction systems. In some cases, this fuel filling station uses the hand cash payment system which is not secured and time-consuming. Each transaction takes more than five minutes to process. For that reason, in some cases, customers face the huge hassle of standing in a long line and waiting for their turn. Not only that, there are high possibilities of fraud activities and robbery being occur for large amounts of the payment transaction. To solve this problem we propose a blockchain-based payment transaction method for fuel filling stations. Here we use the decentralized open ledger infrastructure and proof-of-work to approve each transaction block. Every transaction between the customer and the filling station authority is completed through a digital wallet which is fully secured, fast, and transparent. Comparing to the bank payment transaction system our proposed method is decentralized and has low transaction fees applied in every transaction. This transaction process is free from third-party involvement and all transactions are immutable. For that reason no issues of customer trust and safe from fraud activities in a large number of payment transactions. Our proposed payment transaction method can play an important part to handle large amounts of transactions and provide transaction security for increasing the number of fuel filling stations in South Asia's most populated country.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117319413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10055366
Sadman Jahin, Md Moniruzzaman, Fahmeed Mahmud Alvee, Inzamum Ul Haque, K. Kalpoma
In this work, first, we created an electronic stethoscope (e-Stethoscope) of very low cost that converts the acoustic sound waves obtained through the chest piece into electrical signals and can amplify heart murmurs and noises created by the heart valves. This paper presents an effective way of predicting heart diseases based on heart sounds produced by this e-stethoscope. Our prediction system collects heart sounds from patients using this e-stethoscope and then analyzes them to predict the disease by running various Machine-learning and Deep-learning models like KNN, SVM, Decision Tree, Random Forest, MLP Classifier, ANN, 1D CNN, 2D CNN, etc. We analyzed the results through the 3 datasets, Physionet, Pascal, and Our Collected Heart Dataset. MLP classifier and ANN both performed well on our dataset. A modern heart sound database platform is developed to impact the telemedicine sector worldwide. This telemedicine service may help to cut costs and travel time massively.
{"title":"A Modern Approach to AI Assistant for Heart Disease Detection by Heart Sound through created e-Stethoscope","authors":"Sadman Jahin, Md Moniruzzaman, Fahmeed Mahmud Alvee, Inzamum Ul Haque, K. Kalpoma","doi":"10.1109/ICCIT57492.2022.10055366","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055366","url":null,"abstract":"In this work, first, we created an electronic stethoscope (e-Stethoscope) of very low cost that converts the acoustic sound waves obtained through the chest piece into electrical signals and can amplify heart murmurs and noises created by the heart valves. This paper presents an effective way of predicting heart diseases based on heart sounds produced by this e-stethoscope. Our prediction system collects heart sounds from patients using this e-stethoscope and then analyzes them to predict the disease by running various Machine-learning and Deep-learning models like KNN, SVM, Decision Tree, Random Forest, MLP Classifier, ANN, 1D CNN, 2D CNN, etc. We analyzed the results through the 3 datasets, Physionet, Pascal, and Our Collected Heart Dataset. MLP classifier and ANN both performed well on our dataset. A modern heart sound database platform is developed to impact the telemedicine sector worldwide. This telemedicine service may help to cut costs and travel time massively.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"124 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115882613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10055902
Senjuti Rahman, M. Hasan, A. K. Sarkar
Heart plays a crucial role in all forms of life. Heart-related disorders demand higher precision, consistency, and accuracy in diagnosis and prognosis because even a small mistake might lead to death. Heart-related deaths are common, and the number of these deaths is rising rapidly day by day. Heart disease (HD) prediction with an acceptable level of accuracy is attainable by using cutting-edge machine learning (ML) and deep learning (DL) algorithms. Making an accurate model using these algorithms can predict and categorize cardiovascular illness with high accuracy and reduce medical testing and human intervention. In this study an assessment between ML and DL was carried out to improve classification models for heart disease prediction based on related performance metrics (Accuracy, Precision, Recall, F-1 score, and AUC curve) using a benchmark dataset from UCI machine learning databases of heart disease. which consists of 14 different heart disease-related features. Extreme Gradient Gradient Boosting (XGBoost), Ada Boost, Light Gradient Boosting Machine, CatBoost, Gradient Boosting, Random Forest, Ridge, Decision Tree, Logistic Regression, K Neighbors, SVM-Linear Kernel, Naive Bayes, and deep neural networks, DNN3(3-layer network) and DNN4(4-layer network) are just a few of the classification models that are successfully used in this work for classification tasks. The highest classification accuracy was attained with the Extreme Gradient Boosting classifier (81.10%) (among the machine learning classifiers). The three layer deep neural network (DNN3) among deep learning approaches has provided the best accuracy of 85.41% when using selected features as input. The gathered results showed that deep neural networks outperformed machine learning techniques.
{"title":"Machine Learning and Deep Neural Network Techniques for Heart Disease Prediction","authors":"Senjuti Rahman, M. Hasan, A. K. Sarkar","doi":"10.1109/ICCIT57492.2022.10055902","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055902","url":null,"abstract":"Heart plays a crucial role in all forms of life. Heart-related disorders demand higher precision, consistency, and accuracy in diagnosis and prognosis because even a small mistake might lead to death. Heart-related deaths are common, and the number of these deaths is rising rapidly day by day. Heart disease (HD) prediction with an acceptable level of accuracy is attainable by using cutting-edge machine learning (ML) and deep learning (DL) algorithms. Making an accurate model using these algorithms can predict and categorize cardiovascular illness with high accuracy and reduce medical testing and human intervention. In this study an assessment between ML and DL was carried out to improve classification models for heart disease prediction based on related performance metrics (Accuracy, Precision, Recall, F-1 score, and AUC curve) using a benchmark dataset from UCI machine learning databases of heart disease. which consists of 14 different heart disease-related features. Extreme Gradient Gradient Boosting (XGBoost), Ada Boost, Light Gradient Boosting Machine, CatBoost, Gradient Boosting, Random Forest, Ridge, Decision Tree, Logistic Regression, K Neighbors, SVM-Linear Kernel, Naive Bayes, and deep neural networks, DNN3(3-layer network) and DNN4(4-layer network) are just a few of the classification models that are successfully used in this work for classification tasks. The highest classification accuracy was attained with the Extreme Gradient Boosting classifier (81.10%) (among the machine learning classifiers). The three layer deep neural network (DNN3) among deep learning approaches has provided the best accuracy of 85.41% when using selected features as input. The gathered results showed that deep neural networks outperformed machine learning techniques.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123254100","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}
Properly taking care of us becomes difficult when there is a risk of spreading disease while receiving health care, and the health of many others is threatened by this type of pandemic situation. If a project is designed to avoid such a situation, it can perform the necessary steps for first aid without human contact, such as automatically sanitizing and checking the patient's oxygen saturation level, heart rate or temperature measurement and be able to provide this service to many people at a time without a man-to-man contact. To implement this prototype project, line-following the IR sensor and creating its movement step with fuzzy logic. BPM, SpO2, and temperature sensors are utilized to take data from the patient. All data is processed in NodeMCU, and it’s shown to a web server or app through the Internet of Things (IoT). With its autonomous management system, many service recipients will benefit from it at home or in the hospital. As a result, they can use IoT to monitor their current health state and condition. All the data is stored on the server, allowing any decision-making to play an effective role as the patient's history is known even during the next treatment. However, this reduces the chance of disease spreading and allows many patients to complete the steps before receiving their demanding services.
{"title":"Fuzzy Logic Controlled an Autonomous Patient's Health Monitoring System through the Internet of Things","authors":"Thohidul Islam, Md. Jasim Uddin Qureshi, Md. Farhan Nasir, R. Chowdhury, Hrishin Palit, Papri Mitra","doi":"10.1109/ICCIT57492.2022.10055115","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055115","url":null,"abstract":"Properly taking care of us becomes difficult when there is a risk of spreading disease while receiving health care, and the health of many others is threatened by this type of pandemic situation. If a project is designed to avoid such a situation, it can perform the necessary steps for first aid without human contact, such as automatically sanitizing and checking the patient's oxygen saturation level, heart rate or temperature measurement and be able to provide this service to many people at a time without a man-to-man contact. To implement this prototype project, line-following the IR sensor and creating its movement step with fuzzy logic. BPM, SpO2, and temperature sensors are utilized to take data from the patient. All data is processed in NodeMCU, and it’s shown to a web server or app through the Internet of Things (IoT). With its autonomous management system, many service recipients will benefit from it at home or in the hospital. As a result, they can use IoT to monitor their current health state and condition. All the data is stored on the server, allowing any decision-making to play an effective role as the patient's history is known even during the next treatment. However, this reduces the chance of disease spreading and allows many patients to complete the steps before receiving their demanding services.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125447303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10055778
M. Uddin, Mohammad Khairul Islam, Md. Rakib Hassan, Aysha Siddika Ratna, Farah Jahan
The amount of DNA data is growing exponentially because of enormous applications including gene therapy, new variety development, and evolutionary history tracking. Recently, chaos, kmer count, histogram, and deep learning-based alignment-free (AF) algorithms are widely used for DNA sequence analysis. However, these methods have either high time complexity, memory consumption, or low precision rate. Hence, an optimal solution is needed. Therefore, in this research, a part-wise template matching-based novel similarity feature vector is extracted. Based on this vector, a phylogenetic tree is generated. The method is tested on two benchmark and four standard datasets and compared with recent existing studies. The method achieves 100% accuracy, consumes 10 to 70 times less memory than existing studies, and achieves top-rank benchmark results. Moreover, the required time of this method is very close to the existing best methods. Therefore, in real-time scenarios, industries can use this method with a great level of reliability.
{"title":"A novel part-wise template matching technique for DNA sequence similarity identification","authors":"M. Uddin, Mohammad Khairul Islam, Md. Rakib Hassan, Aysha Siddika Ratna, Farah Jahan","doi":"10.1109/ICCIT57492.2022.10055778","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055778","url":null,"abstract":"The amount of DNA data is growing exponentially because of enormous applications including gene therapy, new variety development, and evolutionary history tracking. Recently, chaos, kmer count, histogram, and deep learning-based alignment-free (AF) algorithms are widely used for DNA sequence analysis. However, these methods have either high time complexity, memory consumption, or low precision rate. Hence, an optimal solution is needed. Therefore, in this research, a part-wise template matching-based novel similarity feature vector is extracted. Based on this vector, a phylogenetic tree is generated. The method is tested on two benchmark and four standard datasets and compared with recent existing studies. The method achieves 100% accuracy, consumes 10 to 70 times less memory than existing studies, and achieves top-rank benchmark results. Moreover, the required time of this method is very close to the existing best methods. Therefore, in real-time scenarios, industries can use this method with a great level of reliability.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125508594","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10056094
P. Das, Nurul A. Asif, M. Hasan, S. H. Abhi, Mehtar Jahin Tatha, Swarnali Deb Bristi
Nowadays, our home is designed with various technologies which have increased our living comfort and offering more flexibility. Installing various technology in our Home makes it a smart home and we also call this installation process Home Automation. The popularity of Home Automation systems is increasing rapidly and it develops the quality of living. Home automation offers automatic light, fan, temperature, etc. control and also an automatic alarming system to alert the people, etc. Already there are various techniques have been used for implementing Home Automation. Here, in this paper, an intelligent door controller, an application of home automation is presented by using deep learning techniques. An intelligent door basically opens automatically and closes after a predefined time based on the person coming in front of the door. If a person is known then the door will be opened and after his/her entrance the door will be closed automatically. And if the person is not known then the door will remain closed. Here to identify the person, the person’s face is recognized by using deep learning. As well ass, Arduino and Servo motors are used to control the door opening or closing.
{"title":"Intelligent Door Controller Using Deep Learning-Based Network Pruned Face Recognition","authors":"P. Das, Nurul A. Asif, M. Hasan, S. H. Abhi, Mehtar Jahin Tatha, Swarnali Deb Bristi","doi":"10.1109/ICCIT57492.2022.10056094","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10056094","url":null,"abstract":"Nowadays, our home is designed with various technologies which have increased our living comfort and offering more flexibility. Installing various technology in our Home makes it a smart home and we also call this installation process Home Automation. The popularity of Home Automation systems is increasing rapidly and it develops the quality of living. Home automation offers automatic light, fan, temperature, etc. control and also an automatic alarming system to alert the people, etc. Already there are various techniques have been used for implementing Home Automation. Here, in this paper, an intelligent door controller, an application of home automation is presented by using deep learning techniques. An intelligent door basically opens automatically and closes after a predefined time based on the person coming in front of the door. If a person is known then the door will be opened and after his/her entrance the door will be closed automatically. And if the person is not known then the door will remain closed. Here to identify the person, the person’s face is recognized by using deep learning. As well ass, Arduino and Servo motors are used to control the door opening or closing.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128637742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-17DOI: 10.1109/ICCIT57492.2022.10055090
S. M. Taslim Uddin Raju, M. Hashem
Diabetes is a perpetual metabolic issue that can prompt severe complications. Blood glucose level (BGL) is usually monitored by collecting a blood sample and assessing the results. This type of measurement is extremely unpleasant and inconvenient for the patient, who must undergo it frequently. This paper proposes a novel real-time, non-invasive technique for estimating BGL with smartphone photoplethysmogram (PPG) signal extracted from fingertip video and deep neural networks (DNN). Fingertip videos are collected from 93 subjects using a smartphone camera and a lighting source, and subsequently the frames are converted into PPG signal. The PPG signals have been preprocessed with Butterworth bandpass filter to eliminate high frequency noise, and motion artifact. Therefore, there are 34 features that are derived from the PPG signal and its derivatives and Fourier transformed form. In addition, age and gender are also included as features due to their considerable influence on glucose. Maximal information coefficient (MIC) feature selection technique has been applied for selecting the best feature set for obtaining good accuracy. Finally, the DNN model has been established to determine BGL non-invasively. DNN model along with the MIC feature selection technique outperformed in estimating BGL with the coefficient of determination (R2) of 0.96, implying a good relationship between glucose level and selected features. The results of the experiments suggest that the proposed method can be used clinically to determine BGL without drawing blood.
{"title":"DNN Based Blood Glucose Level Estimation Using PPG Characteristic Features of Smartphone Videos","authors":"S. M. Taslim Uddin Raju, M. Hashem","doi":"10.1109/ICCIT57492.2022.10055090","DOIUrl":"https://doi.org/10.1109/ICCIT57492.2022.10055090","url":null,"abstract":"Diabetes is a perpetual metabolic issue that can prompt severe complications. Blood glucose level (BGL) is usually monitored by collecting a blood sample and assessing the results. This type of measurement is extremely unpleasant and inconvenient for the patient, who must undergo it frequently. This paper proposes a novel real-time, non-invasive technique for estimating BGL with smartphone photoplethysmogram (PPG) signal extracted from fingertip video and deep neural networks (DNN). Fingertip videos are collected from 93 subjects using a smartphone camera and a lighting source, and subsequently the frames are converted into PPG signal. The PPG signals have been preprocessed with Butterworth bandpass filter to eliminate high frequency noise, and motion artifact. Therefore, there are 34 features that are derived from the PPG signal and its derivatives and Fourier transformed form. In addition, age and gender are also included as features due to their considerable influence on glucose. Maximal information coefficient (MIC) feature selection technique has been applied for selecting the best feature set for obtaining good accuracy. Finally, the DNN model has been established to determine BGL non-invasively. DNN model along with the MIC feature selection technique outperformed in estimating BGL with the coefficient of determination (R2) of 0.96, implying a good relationship between glucose level and selected features. The results of the experiments suggest that the proposed method can be used clinically to determine BGL without drawing blood.","PeriodicalId":255498,"journal":{"name":"2022 25th International Conference on Computer and Information Technology (ICCIT)","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130957749","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：