http://103.15.140.189/handle/123456789/136 2026-04-05T18:04:29Z http://103.15.140.189/handle/123456789/275 Comparative Analysis on Deep Learning Methods in Detecting Pneumonia and Tuberculosis from Chest X-Ray Munia, Shifat Jahan; dash, Tushe; Ritu, Nowshin Tabassum; tasnim, Nishat; Karim, Md Shafiaz Tuberculosis and pneumonia are debilitating respiratory diseases that pose significant health challenges globally. Timely and accurate diagnosis of these diseases is crucial for effective treatment and control. This has led to the development of innovative technologies for the detection and diagnosis of tuberculosis and pneumonia, including the use of deep learning models and computer-aided diagnosis systems. In this study, we have curated a comprehensive dataset containing chest X-ray images with four distinct classes: normal, tuberculosis-infected, pneumonia-infected, and co-infected cases. We have implemented deep learning models, including InceptionV3 classify these images. Our results show that InceptionV3 achieved the highest accuracy of 98.76 percent in disease classification. This system holds the potential to assist healthcare professionals in the early and accurate diagnosis of tuberculosis and pneumonia, ultimately improving patient outcomes and reducing the burden on healthcare systems. Internship Report 2023-11-01T00:00:00Z http://103.15.140.189/handle/123456789/274 Cardiovascular Disease Prediction with Machine Learning Zahan, Md Sarwar; Khatun, Naziya; Hossain, Md Al Sharif; Tasnim, Ramisa; Ahmed, Sabbir Cardiovascular disease (CVD) is a significant global health concern and a leading cause of death. A machine learning odel is the best predictor for the early detection and accurate prediction and prevention of CVD and the best prospects for the study. We begin by creating and segmenting datasets rich in health-related characteristics carefully. We play an essential role in preserving accuracy with data processing models that power data transformation, feature selection, and eliminating outliers. By verifying these steps, the dataset is carefully prepared for predictive modelling purposes. Various models such as XGBoost, Logistic Regression, LightGBM, K-Nearest Neighbors, Gaussian Naive Bayes, Random Forest, Decision Tree, Extra Tree, AdaBoost, Gradient Boosting, Support Vector Machine, and CatBoost are tested before searching for CVD. Our Analysis refers to teaching and testing accuracy to determine the model's performance and generalization ability. CatBoost Classifier has been recognized as an expert performer, demonstrating exceptional test accuracy [1] and literacy when applied to unfamiliar data. However, we analyze the essential features and provide valuable insight into factors in CVD prediction, which positively influence the prognosis of CVD disease. If CatBoost exhibits state-of-the-art accuracy levels, the hyperparameter tuning [2] offers more capability, representing a promising avenue for future research efforts. In summary, this study summarizes, augments, and improves the application of machine learning in cardiovascular disease prediction. These results are not only useful for healthcare practitioners and researchers but also underscore the importance of AI, data processing, and feature selection in healthcare analytics. This work establishes a solid foundation for future research efforts, which join the advancement of medical science and the exploration of healthcare avenues. Internship Report 2023-11-01T00:00:00Z http://103.15.140.189/handle/123456789/273 Online Shopping System Ahmed, Raju; Hasan, Md. Mehadi; Rahman, Md. Mushfikur; Rahaman, Mizanur; Islam, Md. Shariful Ecommerce is an option to buying and selling products via internet. There are so many people who go to market and buy their needy products physically and after that they back home by their own or public transport. But the problem is they waste their valuable time. That’s why working with E-commerce for the people who have less time. E-commerce makes our life easy and less waste time. Customer can choose their products from the site and compare the price and products quality. Firstly, make sure that, the product quality is good and secure the system. In Bangladesh, there are so many e-commerce sites that they don’t fulfill the customer service. Actually, they are not secure. After order the products customer need to get this with less time, but at present some e-commerce company say, they will give it will less time but after order they take so much time. Which is untreatable. At this time, make sure that application is trust worthy. The e-commerce company need to secure and trustable the services and delivers the products with less time. And take the advantage from the customer. Customer wants their products good and get early. Internship Report 2023-11-01T00:00:00Z http://103.15.140.189/handle/123456789/272 IOT Based Automated Gate Unlocking System Emon, Asadullah; Rahman, Tanvir; Mahmud, Md. Rakib; Rahman, Ashikur; Ayatullah, Md. The gate unlock system using Arduino Uno, servo motor, and ultrasonic sensor is an automated system that enables the unlocking and locking of a gate without any human intervention. This project aims to provide a simple and efficient solution to the problem of gate locking and unlocking. The system uses an Arduino Uno, which is a microcontroller board, to control the operation of the servo motor and ultrasonic sensor. The servo motor is used to control the locking and unlocking of the gate, while the ultrasonic sensor detects the distance of the person or object from the gate. The working principle of the system is as follows: when a person approaches the gate, the ultrasonic sensor detects the distance between the person and the gate. If the person is within the specified range, the servo motor unlocks the gate. The person can then enter or exit, and once they are out of range of the ultrasonic sensor, the servo motor locks the gate again. This project has several advantages over traditional gate unlocking systems. Firstly, it eliminates the need for manual operation, which can be inconvenient and time-consuming. Secondly, it provides enhanced security as the gate can only be unlocked if the person is within the specified range of the ultrasonic sensor. Finally, it can be easily customized and adapted to suit specific needs. The gate unlock system using Arduino Uno, servo motor, and ultrasonic sensor has potential applications in areas such as home automation, security systems, and access control. It can be used in residential and commercial properties, and can be easily installed and maintained by anyone with basic electronics knowledge. Overall, this project is an innovative and cost-effective solution to the problem of gate unlocking and locking. Internship Report 2023-07-21T00:00:00Z