Browsing by Author "Alvi, Syada Tasmia"

Now showing 1 - 10 of 10

A Framework for Covid-19 Vaccine Management System Using Blockchain Technology
(Daffodil International University, 2022-06-02) Mahamud, Shafayet; Alvi, Syada Tasmia
The SARS-CoV-2 first surfaced in 2019 in China and later spread across the globe causing a pandemic. Immunisation has thus far been considered to be mankind’s weapon of choice in the frontline fight against the virus defined as Covid-19. Mass vaccination programmes carried out by nations are closely related to public health information, data safety and data security. As countries roll out the immunisation efforts, cyber offenders try to exploit people’s personal health data and inoculation records while citizens can also be exposed to fake vaccine certificates issued by hackers. To prevent such any data breach or data exploitation, an effective system is urgently required to be in place that ensures the maximum security at a time of the unprecedented global crisis. Blockchain can be the perfect solution in this case thanks to its transparency, trustworthiness, and decentralised operations. We have proposed a blockchain based framework for covid 19 vaccination process to provide data immutability, transparency and correctness of beneficiary registration for vaccination, eliminating identity thefts and impersonation, tamper proof self-reporting of side effects, person identification and vaccine certification.
A Predictive Analysis of Chronic Kidney Disease by Exploring Important Features
(Daffodil International University, 2022-01-09) Rahman, Mafizur; Islam, Linta; Rana, Masud; Tazim, Malika Zannat; Sorna, Jannatul Ferdous; Alvi, Syada Tasmia
Chronic Kidney Disease is an incurable disease which causes damages to the functions of a kidney gradually. Only proper treatment can prevent the disease from getting worse. Because of proper knowledge about kidney disorders, people had to suffer from this deadly disease. Thus, in this paper, we analyzed certain key features and noticed several interesting relationships with the disease by considering the actual perception of people. We also predict kidney disease by employing various machine learning algorithms including Logistic Regression, Naive Bayes, SVM and KNN. By applying PCA, we observe that there is an improvement in the accuracy for predicting the disease. SVM outperforms other algorithms with 98% accuracy in predicting chronic kidney disease. In future, we will try to find some significant hypothesis that helps us to prevent the disease better.
BSEVOTING
(Scopus, 2021) Alvi, Syada Tasmia; Islam, Linta; Rashme, Tamanna Yesmin; Uddin, Mohammed Nasir
In a democratic society, a citizen's ability to vote is regarded as one of the most significant legal rights he or she may exercise. For e-voting methods, blockchain presents new possibilities to properly meet transparency, integrity, anonymity and many other security properties. As the need for blockchains keeps rising, demand for bigger, more scalable, more adaptable, and more cost-effective multipurpose chain is also high. Conventional blockchains are incapable of meeting all of these demands. To solve the problems (mostly performance) associated with main blockchains, sidechain technology has recently evolved as a separate chain connected to the main chain that runs in parallel with transactions. Our proposed method is designed to operate on a public blockchain, but we separate the storage of voting information of each candidate using a sidechain to offer a cost-effective blockchain-based voting mechanism by ensuring the security properties such as anonymity, integrity, privacy, security, fairness, receipt freeness and so many. In future, we will broadly discuss and implement this voting system.
Design and Numerical Analysis of a Novel Photonic Crystal Fiber Based Chemicals Sensor in the THz Regime
(Elsevier, 2022-07-22) Hossain, Md Selim; Hossen, Rakib; Alvi, Syada Tasmia; Sen, Shuvo; Al-Amin, Md.; Hossain, Md. Mahabub
We presented a decagonal cladding and hexahedron core-based photonic crystal fiber (PCF) to sense chemicals in the terahertz frequency (THz). Circular air holes (CAHs) in the cladding region make up the proposed sensor. A wide variety of frequencies were evaluated to analyze the sensor’s performance in terms of sensitivity, confinement loss, and effective material loss respectively. We designed and quantitatively analyzed the optical properties of our proposed hexahedron-based PCF sensor using the finite element method (FEM). Square-shaped air hole length, strut, and core size have also been researched to improve the performance of the proposed sensor’s sensing components and fabrication tolerance. At ideal conditions, the suggested PCF sensor has a maximum relative sensitivity of 94.65%, confinement loss of 6.01 × 10− 8 cm− 1 , effective material loss (EML) of 9.16 × 10− 4 cm− 1 , and effective mode area (EMA) of 1.35 × 10− 7 m2 . We are confident that the suggested sensor’s optimized geometrical structure will be manufacturing-friendly, as well as the sensor’s contribution to practical uses. Furthermore, our proposed PCF fiber will be ideal in the terahertz (THz) regions for various optical communication applications and medicinal signals.
Design of Quasi-Shaped Spectroscopy Based Optical Sensor for the Detection of Alcohol
(Elsevier, 2023-08-29) Hossain, Md. Selim; Hossen, Rakib; Walid, Md. Abul Ala; Alvi, Syada Tasmia; Al-Amin, Md.; Sen, Shuvo; Azad, Mir Mohammad
In this research, an entirely novel optical sensor for alcohol detection based on quasi-shaped spectroscopy is offered. A quasi shape with a tightly bonded structure in the cladding region and a hexahedron core region makes up the recommended design, which has a remarkable sensitivity of about 91.35%, 92.55%, and 90.40%and low confinement losses (CLs) of 5.44 × 10−08 dB/m, 6.75 × 10−08 dB/m, and 5.85 × 10−08 dB/m for detecting alcohol like ethanol (n = 1.354), butanol (n = 1.3993), and propanol(n = 1.384), at 1 THz. The numerical aperture, the effective area, and the V-parameter have all been identified and assessed. The 0.8THz to 3 THz operational wavelength range is stated. In the COMSOL Multiphysics (Version 5.6) environment, the properties of these suggested alcohol sensors are numerically examined utilizing the FEM (finite element method) with full vectors. In contrast to other efforts, the proposed sensor, which consists of a quasi-lattice PCF with a perfectly matched layer (PML) of a circular air hole shape and TOPAS as the background material, aims to improve sensitivity response. The proposed sensor might be crucial in the alcohol detection process due to its superior sensitivity response, a single mode of operation across the entire operational terahertz range, and shallow confinement loss. So, this PCF can proficiently be pragmatic in many biological sensing and other THz technology domains.
DVTChain
(Daffodil International University, 22-07-01) Alvi, Syada Tasmia; Uddin, Mohammed Nasir; Islam, Linta; Ahamed, Sajib
Voting is a fundamental democratic activity. Many experts believe that paper balloting is the only appropriate method to ensure everyone’s right to vote. But this method is prone to errors and abuse. Many nations utilize digital voting methods to solve the difficulties of paper balloting. A single flaw in digital voting may lead to massive vote-rigging. Election voting methods must be legal, accurate, safe, and convenient. However, issues with digital voting methods may restrict acceptance. Due to its end-to-end verification capabilities, blockchain technology was developed to address these problems. To guarantee We have used blockchain technology anonymity, privacy, verifiability, mobility, integrity, security, and fairness in voting. By using blockchain our proposed system ensures security, privacy, and integrity. This system provides voter anonymity by keeping the voter information as a hash in the blockchain. It also provides fairness by keeping the casted vote encrypted till the ending time of the election. After ending time, the voter can verify their casted vote, ensuring verifiability. To test our protocol, we put it on Ethereum 2.0, a blockchain platform that uses Solidity as a programming language to create smart contracts. The adoption of smart contracts provides a safe means for performing voter verification, ensuring the correctness of voting results, making the counting system public, and protecting against fraudulent activities. We analyzed the system’s performance based on security and gas costs. It improves in terms of security characteristics and the related cost for the necessary infrastructure.
Improving the Accuracy of Heart Disease Prediction Approach of Machine Learning Algorithms
(IEEE, 2023-05-23) Hossain, Md. Belal; Uddin, Mohammed Nasir; Alvi, Syada Tasmia; Era, Chowdhury Abida Anjum
The work is about forecasting heart disease. First and foremost, we gathered data from various sources and divided it into two portions, one of which is 80% and the other is 20%, where the first part is for training and the remainder is reserved for the test dataset. After collecting this dataset, we applied the pre-processing formula and different classifier algorithms. K-Nearest Neighbor, Support Vector Machine, Decision Tree, Random Forest, Naive Bayes & Logistic Regression are the techniques utilized here. When compared to other algorithms, Logistic Regression, KNN, and SVM provided the same or superior accuracy. Precision, Recall, F1 score, and ERR are used to measure accuracy. Gender, Glycogen, BP, and Heartrate are some of the prefixes used while training and found to be different major vulnerable factors of heart diseases. The direction of this work is real-life experiments and clinical trials using different devices.
Performance Assessment of Advanced CNN and Transformer Architectures in Skin Cancer Detection
(IEEE, 2024-09-26) Sajol, Md Saiful Islam; Alvi, Syada Tasmia; Era, Chowdhury Abida Anjum
Skin cancer has always been one of the most common types of cancer in medical history. Every year, it kills millions of people around the world. To find skin problems early, it is necessary to have a reliable automated method for recognizing them. In the past, protein sequences and different types of imaging methods were used with machine learning to find skin cancer. The problem with machine learning methods is that they need features to be designed by humans, which is hard to do and takes a lot of time. Image processing and deep learning are both used in the method for treating skin cancer that works. Our study is based on The HAM10000 dataset, which is made up of 10015 different dermatoscopic pictures of common pigmented skin lesions from different sources. The newest deep learning techniques, such as Convolutional Neural Networks (CNN), Transformers, and Hybrid models, i.e., ConvNeXt V2, Swinv2, ViT2, Cvt, DenseNet, RestNet, PVTv2, EfficientNet, EfficientFormer, VGG, MobileNet, MobileViTV2, GoogLeNet, are all compared in this study to see which one is better for automatically detecting skin cancer lesions. We found the best model for identifying skin lesions by determining its accuracy, F1-score, Inference time, and confusion matrix. Among our implemented models, the ConvNeXt V2 model has an accuracy of 93.2 % with the lowest inference time. With this research's help, new ways are being made to find skin cancer, which can lead to better patient results and better clinical decisions.
Performance Assessment of Advanced CNN and Transformer Architectures in Skin Cancer Detection
(2024-12-12) Sajol, Md Saiful Islam; Alvi, Syada Tasmia; Abida Anjum Era, Chowdhury
Skin cancer has always been one of the most common types of cancer in medical history. Every year, it kills millions of people around the world. To find skin problems early, it is necessary to have a reliable automated method for recognizing them. In the past, protein sequences and different types of imaging methods were used with machine learning to find skin cancer. The problem with machine learning methods is that they need features to be designed by humans, which is hard to do and takes a lot of time. Image processing and deep learning are both used in the method for treating skin cancer that works. Our study is based on The HAM10000 dataset, which is made up of 10015 different dermatoscopic pictures of common pigmented skin lesions from different sources. The newest deep learning techniques, such as Convolutional Neural Networks (CNN), Transformers, and Hybrid models, i.e., ConvNeXt V2, Swinv2, ViT2, Cvt, DenseNet, RestNet, PVTv2, EfficientNet, EfficientFormer, VGG, MobileNet, MobileViTV2, GoogLeNet, are all compared in this study to see which one is better for automatically detecting skin cancer lesions. We found the best model for identifying skin lesions by determining its accuracy, F1-score, Inference time, and confusion matrix. Among our implemented models, the ConvNeXt V2 model has an accuracy of 93.2 % with the lowest inference time. With this research's help, new ways are being made to find skin cancer, which can lead to better patient results and better clinical decisions.
Short Term Weather Forecasting Comparison Based on Machine Learning Algorithms
(IEEE, 2023-08-24) Era, Chowdhury Abida Anjum; Rahman, Mahmudur; Alvi, Syada Tasmia
Forecasting is the term used to describe the attempt to predict outcomes in unknown or uncertain situations. The most vital factor in many applications of weather forecasting is air temperature. The air temperature alone can’t be the effecting point of forecasting weather. Moreover, with the advancement of computer technologies, forecasting models have been transformed widely. This paper approached a system that forecasts air temperature using machine learning algorithms. Several regression methods were employed to attempt to predict temperatures. This research evaluated four algorithms (Decision Tree, AdaBoost, Random Forest, and Gradient Boosting) on some meteorological data over three years (2015-2019), where 80 percent of the total data set was utilized for training and tested on 20 percent. The variables used include Wind speed, Relative Humidity, Dew point, and Air pressure. The objective was to determine which regressor achieves better outcomes for forecasting air temperature with the lowest error rate. This research concluded that the Random Forest Regressor is the most accurate in prediction. Here, MAE is used to determine the accuracy. On average, the Random forest had the lowest MAE value of 0.102, which was lower than the outcomes of the other three algorithms.