Browsing by Author "Mohammad Monirujjaman Khan"

Now showing 1 - 17 of 17
  • Thumbnail Image
    Item
    Open Access
    A Web and Mobile Application Based Smart Online Health Care System
    (North South University, 2019) Md TalatMahmud; Md Shahrear Bin Satter; Md Samiul Basir Rabbi; Mohammad Monirujjaman Khan; 1511653042; 1210128042; 1511703042
    This report presents the design and the implementation of a web and mobile application based system where people can connect with a doctor by video calling those who are in online and patients can be prescribed. Besides that people can also store their all medical information through this system. User of this system will also get notification when they need to take medicine. We have developed a web based system and a mobile application named “Online Health Care”. User of this system expresses the strong desire to own an online web and mobile application which is used for their self-management and directly getting advice from the doctor
  • Thumbnail Image
    Item
    Open Access
    AI-Powered Smart E-commerce Web Application with Blockchain Integration
    (North South University, 2023) Md. Shamim Ferdous; Salman Yousuf; Tasrifa Hossain; Mohammad Monirujjaman Khan; 2013080042; 2012927042; 1731779042
    This report introduces an AI-Powered Smart eCommerce Web Application seamlessly integrated with Blockchain technology. The project aims to enhance the security, personalization, and efficiency of online shopping through the integration of advanced technologies. It is in its completed state, which now has a multilingual AI chatbot, a machine product recommendation engine, and blockchain integration with a mainstream payment processor. The project follows a methodology that includes the use of Python, Django, PostgreSQL, NodeJs for Backend and JavaScript, ReactJS, NextJs for Frontend development. The results include a smart e-commerce platform with superior autonomous customer service, personalized and efficient user experience and safe and transparent payment processing. The web application received an overall grade of “A” from GTmetrix with the largest content paint time of 277 milliseconds. Google’s PageSpeed Insight gave the application a score of 97% for performance and 79% for accessibility. The product recommendation engine has obtained an accuracy of 0.9926. Integration of the blockchain ensured the immutability of the transaction records. The AI chatbot was able to reply to customer queries in English, Bengali, and Romanized Bengali. This project positively impacts society and the economy by creating job opportunities and contributing to the growth of the e-commerce industry.
  • Thumbnail Image
    Item
    Open Access
    Autonomous Air Defense System for National Security
    (North South University, 2020) Fazle Rabby Khan; Md. Muhabullah; Roksana Islam; Mohammad Monirujjaman Khan; 1611089042; 1611090042; 1620686042
    Air defense systems were designed to reduce threats efficiently and thus protect the country from major attacks. It is therefore a fundamental part of any country around the world because it provides national security. This paper presents the development of an Autonomous Air Defense System (AADS) which will automatically detect aerial threats (e.g. drones) and target them without any human intervention. The AADS is implemented using a radar, camera and laser-gun. The Radar system dynamically emits microwaves and detects objects around it. It triggers the camera system if it senses the frequency of any aerial threat. The camera receives the radar’s signal and detects using a neural network whether it’s a threat or not. Neural network algorithms are used for the detection and classification of objects. The Laser gun locks its target if the live video feed classifies an object as a threat of more than 75%. In the detection stage an average loss of 0.184961 was achieved using YOLOv3 and 0.155 using Faster RCNN. This system will ensure that no human errors are made while detecting threats in a region and improve national safety.
  • Thumbnail Image
    Item
    Open Access
    Bangla Hate Speech Detection: Comparative Analysis of Machine Learning Models and Recurrent Neural Networks
    (North South University, 2023) Abdullah Al Shafi; Miraj Hosaain Shawon; Rafiad Rahman Badhon; Mohammad Monirujjaman Khan; 19111515642; 1911329042; 1722044642
    The spread of hate speech on the internet has been related to an increase in violent acts committed against minority groups all across the world. Nowadays, usage of social media is at its peak, and so is hate speech on online social media. Nearly every continent has reported incidents. A majority of the world's population uses Facebook alone, and many people increasingly converse on social media. Although multiple hate speech detection paper has been done based on the Bengali language, there is no paper has been done based on comparative analysis of machine learning and recurrent neural networks regarding Bengali hate speech detection. Therefore, this study aims to train different models that can detect Bengali hate speech on different social media platforms and do a comparative analysis of the models. Bengali hate speech is on the rise on social media platforms, threatening the general public's mental health. Thus, detecting and preventing it from posting on social media is a suitable approach to preventing hate speech from spreading. Using statistical methods, the collected data were analyzed. Also, manually labeled the collected data based on sentiment. To remove punctuation marks, a punctuation remover is used, and regular expressions are used to remove foreign languages from the dataset. Moreover, Bangla natural language toolkit was used to remove Bangla stop words from the data. A label encoding method is used to make the dataset machine readable. Natural language processing toolkit porter stemmer is used for tokenization and for feature extraction, term frequency-inverse document frequency is used for training and testing the dataset, hold-out validation approach was used. Several machine learning and recurrent neural network models, decision tree (DT), K-nearest neighbor (KNN), random forest (RF), support vector machine (SVM), multinomial naïve bayes (MNB), long short term memory (LSTM), bidirectional long short term memory (Bi-LSTM), and convolutional long short term memory (CNN-LSTM) were implemented. Among machine learning models, support vector machine gained 0.96 accuracy, and among recurrent neural network (RNN) models, bidirectional long short term memory gained 0.94 accuracy.
  • Thumbnail Image
    Item
    Open Access
    Design and Investigation of Energy Harvesting System from Noise.
    (North South University, 2021) Junayed Hossain; Nazmus Sadad Ovi; Mohammad Monirujjaman Khan; 1612680643; 1331388043
    Human activity and various activities are needed to live in today's world, and all of these activities create different types of sound. A noise is a noisy or irritating sound that causes disruption, such as street traffic noises, construction sounds, airports, and so on. This paper looks at a less well-known renewable energy source. Using a suitable transducer, noise (sound) energy can be transformed into a viable source of electric power. This can be accomplished by using a transducer to transform noise-induced vibrations into electrical energy. Reducing the strain on the main power grid and reducing fossil fuel imports. A speaker and a transformer are used to convert noise generated by car horns and other noise sources into electrical energy in a proposed application. The theory of electromagnetic induction can be used to transform noise vibrations into electrical energy. A transformer was used to boost the received signal from 0.7 to 2 volts.
  • Thumbnail Image
    Item
    Open Access
    Developing a Mobile Application using Deep Learning for Cataract Classification
    (North South University, 2023) Tasnia Ishrat Khan; Fatima Ibrahim; Mohammad Monirujjaman Khan; 1911539642; 2121340642
    One of the leading global causes of vision loss and blindness is the cataract. The percentage of blind people is around 50%. As a result, early cataract detection and prevention may limit vision loss and blindness. Contrary to cataract, artificial intelligence (AI) has made significant progress in the treatment of glaucoma, macular degeneration, diabetic retinopathy, corneal abnormalities, and age-related eye diseases. However, the vast majority of cataract detection algorithms in use are built using common machine learning techniques. On the other hand, manual extraction of retinal features is a laborious method that needs a skilled ophthalmologist. In order to detect cataracts, we have built the framework of an Android application. We then used algorithms to extract accuracy, graphs, trainable and untrainable parameters, and differentiation of cataract and non-cataract eye images from a gathered dataset. In order to identify the cataract using color fundus images, we presented the VGG19 (Visual Geometry Group), and digital image we presented Inception V3, which is a CNN (convolutional neural network) model. This will be incorporated into an Android application. The results of fundus image, the training procedure demonstrate that the model attained a flawless accuracy of 1.0000 on the training data for epochs 10 to 15. It scored an accuracy of 0.963 on the validation set, which is still quite high. With values ranging from 0.25 to 0.27, the validation loss was similarly largely consistent. The model is doing well and has mastered correctly classifying the photos. On the test data, the model produced a loss of 0.25735 and an accuracy of 0.9241. The result of the digital image, accuracy is 0.973 on the validation set, which is quite high and on the test data, the model produced a loss of 0.26753 and accuracy of 0.93491. The significance of these results is that the model performs effectively, can reliably categorize test photos with high accuracy, and will be trustworthy for patients to utilize.
  • Thumbnail Image
    Item
    Open Access
    Development of An IoT Based Sleep-Apnea Monitoring System
    (North South University, 2023) Hasibul Islam Shanto; Rubaiyat Sharmin; Ilmoon Jahan; Mohammad Monirujjaman Khan; 1912212642; 1911170642; 1911244042
    Sleep apnea is a common and potentially serious sleep disorder that affects an estimated 22 million Americans and is characterized by repeated episodes of shallow or stopped breathing during sleep. These episodes can lead to disrupted sleep and decreased oxygen levels in the blood, which can have serious consequences including high blood pressure, heart disease, stroke, and even death. A real-time sleep monitoring system using Internet of Things (IoT) technology has been developed to detect and alert individuals to sleep disorders such as sleep apnea.The system developed in this study includes sensors that measure various physiological parameters, including electrocardiogram (ECG), heart rate, pulse rate, skin reaction, humidity, temperature, and SpO2, during sleep. These parameters are continuously monitored and any unusual patterns or events are detected and alerted through a mobile application. The mobile application has been developed using MIT app inventor. It shows us the level of the parameters and any abnormality in health can be detected through it. This study is particularly valuable since it can test sleep indices without waking the user and display them in a mobile application at the same time using a Bluetooth module. The system has been designed in such a way that it may be utilized by anyone. Multiple analogue sensors are used in conjunction with the Arduino UNO to measure various sleep factor factors. The method was evaluated and tested on the bodies of many persons. The device watches multiple people as they sleep in order to evaluate and identify sleep apnea in real time. According to our findings, persons with significant heart and lung problems suffer sleep apnea. Sleep apnea is more common in persons over the age of 60. This study also examines those who are not at risk of sleeping disorders based on the data obtained. This document will help everyone learn about sleep apnea and will assist individuals in detecting and preventing it.
  • Thumbnail Image
    Item
    Open Access
    Development of IoT-Based Real-Time Patients Vital Physiological Parameters Monitoring System Using Smart Wearable Sensors
    (North South University, 2022) Ajan Ahmed; Shefat Ara Moon; A.S.M. Sabbir Hassan; Nusrat Jahan Fatima; Mohammad Monirujjaman Khan; 1811222043; 1620561042; 1620182643; 1612202643
    Healthcare is one of the least funded sectors in Bangladesh and many other similar developing countries. There is no health coverage and health insurance is almost non-existent. Thus, people living in rural and remote areas do not have access to proper healthcare and when they do, it is too expensive. The aim of this research was to develop a real-time health monitoring system that is cheap, easy to use and accessible by both doctors and patients. The system consists of several Internet of Things (IoT) based sensors connected to an Arduino microprocessor, which thus measures the body vital signs of the patients. The measured readings are then transmitted to an android application on a smartphone via a Bluetooth module. The sensors are connected to analog inputs. These sensors measure analog data which is amplified by the microprocessor after being sorted. Doctors can also carry out the diagnosis of ailments using the data collected remotely from the patient. An Android based mobile application that interfaces with a web-based application is implemented for efficient patients-doctors dual real-time communication. The android application, which is connected to a mySQL database, updates the said database, which in turn updates and displays the readings on a website accessible by both doctors and patients. The health monitor was initially tested using an Arduino Integrated Development Environment (IDE) monitor and one single user. Once initial simulations were successful, it was tested on a sample size of 5 more patients. In the end, the testing of the wireless health monitor produced successful results that measured patient vitals with a high level of accuracy.
  • Thumbnail Image
    Item
    Open Access
    Effectiveness of Online Classes DuringCovid-19 Pandemic Using Machine Learning
    (North South University, 2021) Asrafi Akter; Naziba Nasir; Saheeb Tareque; Mohammad Monirujjaman Khan; 1711502042; 1612618042; 1721809042
    The deadly COVID-19 started its journey in December 2019 in China. For most of the patients, it’s like a mild fever and non-specific gastrointestinal symptoms to a lesser extent. Aged people with previous illnesses like diabetes, heart problems, and high blood pressure suffer the most. Until 26th May 2021, there have been 169,094,726 confirmed cases of COVID-19, including 3,512,510 deaths, reported to the World Health Organization. Since COVID-19, lifestyle has changed for almost everyone. That’s when e-learning had a bigger impact on all the students worldwide. It is no different for the students of Bangladesh, either. To keep the study of the students, the government of Bangladesh made a decision and advised the educational institutions to take classes online. While attending online classes, students are facing many opportunities and obstacles, such as disruption in class, health issues, financial issues, and saving time from traffic. However, being captive, they have utilized their time a lot better, and this has reflected on their performances too. Machine Learning is a sub-region of man-made reasoning, which predicts outcomes depending on the features of a given dataset. In this paper, with the help of a machine learning approach, 6 university students during the pandemic were found. By creating our dataset, the model was trained, and then a prediction was made depending on the different features of the dataset. The collection of data that helped to create a model and gave the highest accuracy over students’ performance. Exploration was done using different algorithms, i.e., Linear Discriminant Analysis (LDA), Logistic Regression, and K-Nearest Neighbor (KNN) for classification. Prediction score Accuracy of 81.05 % by LDA, 86.3% by Logistic Regression, and 84.2% by KNN was achieved. The highest prediction score was achieved by Logistic Regression (86.3%). The accuracy shows the effectiveness of online classes during the pandemic through a Machine Learning approach. It shows how the online classes are effective for a particular student when the input fields are filled. Overall, after performing the three algorithms (K-Nearest Neighbor, Logistic Regression, Linear Discriminant Analysis)with satisfactory results and successfully predicting with a high level of accuracy, while maintaining its objective of being easy to understand.
  • Thumbnail Image
    Item
    Open Access
    Joltorongo A Web based Bengali Music Streaming Platform
    (North South University, 2019) Shirjoy Malek; Ahmed Aziz Nahin; Rifat Hossain; Md. Enamul Haque Imran; Mohammad Monirujjaman Khan; 1511321042; 1520476042; 1520827642; 1520581642
    Music is part of human life from an ancient period and with the technological growth it has become part of our daily life. This project is a music streaming platform which is designed to fulfil all our daily music related needs. Unlike all other music related services in our country, this project will give the taste of fully dynamic and user-friendly service to the lovers of music. It will also help the decline of Bengali music industry. The main problem in the music industry now-a-days is to let the music lovers know what songs and albums are coming out soon so sells don’t reach expectation. This product aims to solve this huge gap in communication by adding new released tab which would feature the new albums that are released. Another major problem is the hardship of the up and coming artists who don’t get enough recognition. This project will give them platform to show off their talent to the wider audience through the use of new artists tab so they can send their songs to the masses without spending huge amount of cash. This project would use amazon web server to store the songs so there would be no cases of down servers or security issues and encryption would be used to confront piracy. Furthermore, this project aims to be the go-to place for the Bengali music lovers and music lovers like to keep in touch with music from all round the globe. This project would allow browsing top music news sites while listening to songs to keep in touch with all thing’s music. Overall this project wants to provide a platform by which Bengali music industry can thrive once more
  • Thumbnail Image
    Item
    Open Access
    KenaBecha: An E-Commerce Website And Mobile Application For Buyers And Sellers
    (North South University, 2018) Irtiza Nahian Faiza Sanha; Md.Shah Alam Chowdhury; Shamanna Ferdous Haque; Mohammad Monirujjaman Khan; 1520448042; 1430501042; 1410904042
    With growing use of internet, people are now able to reach out to the world as fast and as efficiently ever. In this project, an interactive multimedia is developed where sellers from rural areas will be able to contact with the buyers all around the country directly. The system will be accessible via internet and available to the target population through computers or laptops and even smart phones very easily. The project includes content designing, web portal development and application development. We hope that the project will make lives easier for the targeted population. The buyers and sellers will be able to view the products online. This would eliminate any middlemen among them and make it much easier for users in terms of usability
  • Thumbnail Image
    Item
    Open Access
    Product Recommendation System by Analyzing Customer Reviews and implementation using Django Based E-commerce website
    (North South University, 2021) KAYSER AHMED; SUMIAYA KHAN; MD. AMZAD HOSSEN; Mohammad Monirujjaman Khan; 1620252042; 1711879642; 1620396642
    This report presents a recommendation system and its implementation on a real time Ecommerce website. This will help the customer to discover new products that would please a customer, recommend products by monitoring customers activity like searching of products, along with the previous set of things a user enjoyed and also find a selection of products that would be enjoyed by a community of people. The biggest secret of achieving success in business is to provide better quality customer services that ensures customer satisfaction. A statistic shows that about 70% of purchase decisions are taken relying on how consumers believe, they are being served. Online shopping is on trend now a days where customer buys things or services without any middle way. However, to give customer satisfaction and earn more revenue, number of advertising ways are introduced. One of which is intelligently recommendation services or products. A product recommender system is a software that detects the consumer’s behaviour on e-commerce sites and on the basis of that, suggests products that meets interest of consumers. This paper presents the design and implementation of a recommendation system on Ecommerce which will suggest the consumer with their relevant and desired products and make their online shopping more comfortable. For implementing this system using machine learning, three algorithms (Model Based Collaborative Filtering, Popularity Based Filtering, K means) are used. The system is designed in such a way that focused on a fresh customer's first visit on the company's website to when they perform repeat purchases. The system will be proved as user friendly and will effectively predict which products customers would like the most. The accuracy in matrix factorization using SVD is 97.7 percent and with KNN using item similarity is 84.56 percent.
  • Thumbnail Image
    Item
    Open Access
    Smart Detection of Deepfake Using Various Deep Learning Models
    (North South University, 2023) Mariam Binte Bashir; Iftekher Mahbub Rafi; Mohammad Monirujjaman Khan; 2021874042; 2021463642
    Deepfake is a digitally manipulated image or video that is generated using an algorithm to replace the original person with someone else which looks authentic. The need to create efficient techniques for identifying and categorizing counterfeit photos on digital platforms has become crucial due to their widespread presence. This project does the prediction of counterfeit photos by employing of deep learning (DL) and transfer learning methodologies, specifically leveraging convolutional neural networks (CNNs) models. Our work focuses to enhance the accuracy and reliably detecting deep fake images. The DL models have been trained using the dataset acquired from Kaggle, titled "140k Real and Fake Images". Transfer learning techniques are employed to enhance prediction performance by leveraging knowledge from large datasets and applying it to pre-trained models. Evaluation criteria such as accuracy, precision, recall, and F1 score are employed to assess the ability of a model to effectively distinguish between genuine and manipulated images. The models exhibit strong discriminatory capabilities and provide reliable picture classifications. The study investigates the efficacy of various deep-learning models in the task of image classification. The Vgg16 achieved a peak accuracy of 99%, showcasing the possibilities of deep learning algorithms. MobileNetV3s and DenseNet121 demonstrate their efficacy in faulty image categorization by achieving accuracies of 98%, 94%, and 95% correspondingly. Unlike other models, Xception achieves remarkable accuracy rates of 96% through training on a wide range of datasets containing both genuine and manipulated images. This enables it to effectively differentiate between real image and deepfakes, which we deployed to build a website. Thus, our experiment demonstrates the versatility of deep learning models in handling various image formats.
  • Thumbnail Image
    Item
    Open Access
    The Prediction of Stock Market Using Recurrent Neural Network
    (North South University, 2021) Sadman Bin Islam; Mohammad Mahabubul Hasan; Mohammad Monirujjaman Khan; 1611957042; 1421274042
    Stock price forecasting is becoming increasingly popular recently in the financial realm. Shares price prediction is important for increasing the interest of speculators in putting money in a company's stock in order to grow the number of shareholders in the stock. Successfully predicting the future price of a stock could result in a sizable return. When it involves forecasting, various methodologies are used. This report uses a replacement stock price prediction framework is proposed utilizing a well-liked model which is Recurrent Neural Network (RNN) model i.e., Long Short-Term Memory (LSTM) model. It is often shown from the simulation results that utilizing these RNN models, i.e., LSTM, and with proper hyper-parameter tuning, the proposed scheme can forecast future stock trend with high accuracy. The RMSE for LSTM model was measured by varying the number of epochs, difference between predicted stock price and actual stock price. The model is trained and evaluated for accuracy with various sizes of knowledge. The assessments are conducted by utilizing a freely accessible dataset for stock markets having date, volume, open, high, low, and closing prices. The major goal of this article is to determine to what degree a Machine Learning algorithm can anticipate the stock market price with greater accuracy.
  • Thumbnail Image
    Item
    Open Access
    Travel Buddy Finder
    (North South University, 2019) Md.Farhan bin shafiq; Md.Shamsuzzaman; Md.Imam hossain; Mohammad Monirujjaman Khan; 1411464042; 0930037042; 1410242042
    In this project, we develop and design an online-based application. Travel Buddy is a FREE social networking app that helps you find travel buddies and partners for your backpacking or holiday plans. We developed this application to automate a travel agency. This client-server-based app will serve all the agents of a travel agency and record all the information of a passenger. The agency can discover all the money transactions and passenger services from remote places through this platform. This social networking app also supplies necessary reports. In this project, there is a login page for each travel agency employee. The admin user can define roles for all users with respect to their user type. That means the admin can authenticate which kind of user can see or perform operations on which pages. This online-based app is device-independent.
  • Thumbnail Image
    Item
    Open Access
    Vhromon - A Online Hotel and Resort Booking System
    (North South University, 2020) Sajjad Mahmud Khan; Sajjad Kashem; S M Imam Jahed Hossain; Mayen Uddin; Mohammad Monirujjaman Khan; 1421222042; 1330273042; 1511913642; 1511500642
    The rise of population and the industrial revolution, it is obvious that hotels and resorts are increasing drastically day by day. In 21st century tourism is a trend. Everyone loves to travel and the reasons vary person to person. Whether it is to have a cheerful and peaceful mind and place or business purposes, either way hotels and resorts actually doing their jobs very well. This, tourism is a promising sector for any country and Bangladesh is no different. But the problem is we have to do the hotel booking and reservation manually most of the time. Sometimes it is too troublesome and knowing about the suitable hotels and resorts is too difficult. Though there are some online platforms available but they are not our country oriented such as the payment system is not preferable for us. To solve this problem we developed Vhromon, a comparison based online based hotel and resort booking system where anyone can book a hotel or resort by just login from the internet. Vhromon is an interactive online platform and it is user friendly and easy to get like the most of the platforms available nowadays. But they do not have all the hotels and resorts enlisted, only the well-known one. On the other hand, Vhromon has everything enlisted where customer can compare them and can choose the right one for them. Furthermore we have a vendor panel where any hotel or resort owner can add their hotel or resort easily without any hassle unlike the existing one.
  • Thumbnail Image
    Item
    Open Access
    Zero-Code Advergame Platform
    (North South University, 2023) Ahnaf Rahat; Nusrat Jahan Riya; Nishat Maharin Bushra; Mohammad Monirujjaman Khan; 1912685042; 1912490642; 1912903642
    The Zero-Code Advergame Platform is a web application where we can automatically develop fully functional advergames. This project’s specialty is that users do not require any prior programming knowledge. We have completed all the core functionalities of a web application. Currently, we have an efficient front-end interface, the essential APIs, an integrated database, a user authentication system, and exceptional features, and among them, “Game Builder” is the most significant one. To build this, we have used Django and JavaScript. We have used the Model-View-ViewModel (MVVM) to promote a transparent separation of concerns, reduce dependencies among components, and enable a more modular and testable application, ultimately leading to a smoother and more efficient development process. We have connected the backend with the front end through API endpoints. APIs provide a systemized way for different software tools to interact with each other, defining regulations to follow while exchanging data. Now, end users can register and access the website securely, and all the listed games are visible and playable. This platform is built specifically for consumers who prefer a hassle-free development environment. To meet this target, we have used React JS and JavaScript to make the user experience easier. In a web application, we must have the flexibility to add new features, and that’s why we have used the Django Rest framework to build this platform. A fully working, sustainable platform with the motto of helping users get the best-developed advertising. It will create enthusiasm for consumers to try this new strategy on their entrepreneurial journey. The Zero-Code Advergame Platform will bring countless benefits and great impacts to their journey without getting ripped off.