Department of Information and Communication Technology (ICT)

Browse

Now showing 1 - 20 of 43

5G Technology and Future in Bangladesh
(Comilla University, 1-Feb-2025) Hasan, Md. Faysal
A Comparative Analysis of Identifying Multi Class Toxic Comment Classification Utilizing Deep Learning and Machine Learning Technologies
(Comilla University, Jan-2025) Sarker, Sajib
Despite extensive efforts to prevent and manage malicious human behavior, there are still many compelling issues in cyber platforms. When such individuals live in a world where technology is rapidly developing and where various online social media platforms like Twitter, Facebook,Instagram, etc. are widely accessible, those nefarious human acts rise. Because of this, extremely dangerous crimes like toxic comments, which pose a threat to users, groups, and even
A Comparative Analysis of Identifying Multi Class Toxic Comment Classification Utilizing Deep Learning and Machine Learning Technologies
(Comilla University, 1-Jan-2025) Sarker, Sajib
Despite extensive efforts to prevent and manage malicious human behavior, there are still many compelling issues in cyber platforms. When such individuals live in a world where technology is rapidly developing and where various online social media platforms like Twitter, Facebook, Instagram, etc. are widely accessible, those nefarious human acts rise. Because of this, extremely dangerous crimes like toxic comments, which pose a threat to users, groups, and even governments, have increased in frequency. This study aims to identify such harmful comments in Bangla text that are posted on social media sites, classify them using machine learning and deep learning algorithms, and take preventative measures. By combining TF-IDF with various machine learning algorithms, such as Naive Bayes, Decision Tree, Random Forest, AdaBoost Classifier, Stochastic Gradient Descent Classifier, Logistic Regression, KNN, and Support Vector Machine, we were able to detect toxic comments with an accuracy of 76.5%, 78.2%, 75.0%, 75.1%, 76.2%, 77.7%, 70.2%, and 78.5%, respectively. In comparison to previous machine learning techniques, we achieved an accuracy of 89% by employing deep learning algorithms
A Deep Learning Model for Classifying Bangladeshi Food and Estimating Calories
(Comilla University, 1-Feb-2025) Nayeem, Md Najmul Islam
Accurate food calorie estimation is essential for managing dietary intake, particularly for
A Project Report on COMILLA UNIVERSITY HALL MANAGEMENT SYSTEM
(Comilla University, 1-Feb-2025) Hasan, Md Mahedi; Rahman, Marfater
Comilla University Hall Management System is a software developed for managing colorful conditioning in the residence. There are five resident halls for the scholars of Comilla University. As the number of halls are adding up, the number of scholars abiding in those caravansaries are also adding. As a result, the people managing the halls are under a lot of stress. Software or applications aren’t generally used in this environment. Generally, in Bangladesh perspective no digital systems or applications are used, most of the halls or residences are managed manually. In early days, managing manually was not very tough. But at present manual system is tedious. This specific design addresses the challenges of hall management and circumvents the issues that arise when carried out by hand. When the drawbacks of the current system are recognized, lead us to the design of a motorized system that will work with the existing system and the system that is additionally easily grasped and more GUI acquainted. We can ameliorate the effectiveness of the system and therefore overcome the downsides of the being system. Comilla University Hall Management System targeted for sale operation of the hall for better control and timely response. This eliminates time day and paper deals being marked. The Warden is handed with a better control over the deals like adding the details of new scholar in the hall, modifying the details of the scholars, deleting the scholars, viewing the pupil’s details abiding in the hall, adding or removing rooms, checking available seats. scholars can also check if there's any room available or not, look for better seats. As the main target of this application is to do efficient management by storing scholar’s data, MongoDB is used as database as it offers numerous benefits, especially its ability to manage unstructured data, its straightforward process for horizontal scaling, and its quicker performance in insert and update tasks, which makes it perfect for applications that require rapidly evolving data structures and handle large volumes of data. In the future, we can implement online fee payments, integrate various load balancers, develop a mobile application, and establish a backup database for improved security. This design’s main aphorism is to reduce the effect of wardens and give better service to the scholars. The main thing of this design is to develop a system for the robotization of the hall affiliated information
A Time-Nonlocal Optimization Approach for Image classification
(Comilla University, 1-Feb-2025) Das, Hridoy Chandra; Hafsa, Akter
Image classification remains a fundamental problem in artificial intelligence, with applications
Academic Assistant :
(Comilla University, 1-Feb-2025) Parves, K. M. Tousif Bin; Islam, Saqiul
The Academic Assistant is an AI-powered robot designed to enhance educational
AI Chatbot For Solving Mathematical Problems Using Large Language Models and Retrieval-Augmented Generation (Rag) With Custom Dataset Integration
(Comilla University, 1-Feb-2025) Ahmed, Junaid; Mahin, Md. Mohiuddin Khan
Mathematical problem-solving is a fundamental skill in education, and recent advancements in artificial intelligence (AI) offer innovative ways to automate this process. This paper presents an AI-powered chatbot developed using Google Vertex AI Agent Builder, designed to solve mathematical problems across various domains. By leveraging Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG), the chatbot integrates a custom dataset to enhance its problem-solving capabilities, allowing it to generate accurate and contextually relevant solutions. The RAG technique enables the model to retrieve external information dynamically, improving the chatbot's ability to handle a wide range of mathematical problems. The integration of a custom dataset ensures that the model can effectively tackle specific problem types, making it adaptable to different scenarios. This work demonstrates the potential of AI-driven tools, powered by Google AI Agent Builder, to provide instant, reliable mathematical assistance, offering valuable support for students and educators alike
An Unified Quantum Classical Model For Noisy Label Medical Image Binary Classification.
(Comilla University, 1-Feb-2025) Bhuiyan, Taki Jakera; Fahim , Jahid Karim
The presence of noisy labels in medical imaging datasets can severely impact diagnostic
Asthma Disease Prediction using Efficient Machine Learning Model
(Comilla University, 1-Feb-2025) Ahmed, Kazi Tanvir; Ali, Md Ahad
Over 200 million people worldwide suffer from asthma, a common chronic respiratory condition that claims about 450,000 lives each year. The application of machine learning to aid in decision-making in the healthcare industry is growing. It works especially well for activities like diagnosis, prediction, and clinical insight in the management of asthma. By seeing trends, predicting future asthma episodes, and supporting individualized care regimens, these technologies assist enhance treatment results. By adjusting hyperparameters and implementing effective models, this work aims to enhance the prediction of asthma disease through machine learning algorithms. Many machine learning models, including XGBoost, Random Forest, K-Nearest Neighbors (kNN), Decision Tree, and Logistic Regression, were trained and evaluated using a dataset of 316801 unbalanced samples from Kaggle, enhanced by 285 survey replies. With an accuracy of 0.7012 and superior performance over other models in terms of precision, recall, and F1-score, XGBoost stood out among the rest. Model dependability was increased by applying SMOTETomek to overcome data imbalance. In order to improve classification accuracy, the study also investigates the effects of several feature selection strategies and hyperparameter tweaking. According to the results, using machine learning models that have been tuned can greatly enhance asthma prediction, enabling early diagnosis and treatment. The effort will concentrate on creating a real-time forecasting system for clinical applications, and taking genetic and environmental factors into account. This study advances AI-powered medical solutions for the diagnosis of asthma. According to this study, machine learning (ML) has the potential to be a very effective technique for forecasting asthma flare-ups. Nevertheless, these models' approaches varied widely. Future research should concentrate on enhancing the models' practicability and generalizability in order to promote their use in clinical practice
Automatic Detection of Flood Severity Level from Flood Videos using Deep Learning Models
(Comilla University, 1-Feb-2025) Hossen, Akther
Floods are one of the most destructive natural disasters, causing significant loss of life, property, and infrastructure. Accurate and timely assessment of flood severity levels is crucial for effective disaster management and mitigation. This project focuses on the automatic detection of flood severity levels from real-time flood videos using deep learning models. Leveraging advancements in computer vision, the proposed system extracts meaningful features from video frames to classify
BRAIN STROKE PREDICTION USING MACHINE LEARNING
(Comilla University, 1-Feb-2025) Sarkar, Md. Mahmudur Rahman; Sarkar , Proma
The present document is a project carried out in the field of engineering to create a system
Classifying Spam Email Using Machine Learning
(Comilla University, 1-Feb-2025) Sana, Priyosree
Email is a widely used method of transferring data and information via digital devices. Email is widely used for data communication between servers by millions of users worldwide. Unwanted emails, or spam, have become a problem for big corporations and organizations as the number of spam grows rapidly each year. Spam is both unpleasant and perhaps destructive. Spam consumes computer, server, and network resources, creating bottlenecks and slowing down digital devices. Furthermore, consumers spend a significant amount of time deleting unsolicited emails. Spam emails are a huge cybersecurity concern, with research indicating that 91% of internet assaults begin from malicious emails, while 97% of users fail to accurately detect them. To reduce these dangers, this work proposes a machine learning-based solution to spam email categorization. The dataset is preprocessed with Natural Language Processing (NLP) techniques before features are extracted using the Term Frequency-Inverse Document Frequency (TF-IDF) vectorization approach. To find the best machine learning model for spam detection, five models are trained and assessed: Logistic Regression, Random Forest Classifier, Decision Tree Classifier, K-Nearest Neighbours Classifier, Multi-Layer Perceptron (MLP) Classifier. The models are compared using performance criteria such accuracy, precision, recall and F1-score. The findings show that it has the best accuracy, which makes it a dependable automatic spam filtering solution. By offering an effective spam categorisation model that can shield users from phishing and other online dangers, this study helps to improve email security.
CNN-Based Vehicle Rules Violation Detection System
(Comilla University, 1-Feb-2025) Khasrul Alam, Khasrul
One of the most well-known components of the cutting-edge ai and machine learning sector that tracks numerous functional activities and their influence on the growth of modern smart cities worldwide is the video based intelligent transportation system v-its these days the high accident rate in a crowded metropolis is a major concern it is now crucial to address all of the causes of the significant increase in traffic accidents and the percentage of fatalities that are frequently attributed to breaking traffic laws. An enhanced V-ITS system for identifying and classifying cars, drivers, and their license plates while travelling on highways has been developed in order to overcome those difficulties. The technique is reliable for reducing the rate of violations because the upgraded V-ITS system can perform the aforementioned tasks with precision from real-time traffic footage.
Comparative Analysis of Fruits Disease Detection using Different types of CNN Architecture
(Comilla University, 1-Feb-2025) Islam, Md. Tohidul; Alam, Kursed
Fruit diseases faces significant challenges to agricultural productivity, leading to substantial economic losses globally. Early and accurate detection of these diseases is crucial for effective management and prevention. In recent years, Deep Learning, particularly Convolutional Neural Networks (CNNs), has emerged as a powerful tool for image classification and disease detection in agriculture.
Development of a Healthcare Chatbot System
(Comilla University, 1-Feb-2025) Rahman, Md. Tareq
In order to live a better life, healthcare is crucial. Technology has never played a more importan
Diabetes Prediction using Machine Learning
(Comilla University, 1-Feb-2025) Pal, Prokash; Tajrin, Nushera
Particularly in countries like Bangladesh, where delayed diagnosis and limited healthcare resources exacerbate its impact. Traditional diagnostic methods are often costly and time-consuming, leading to late-stage detection and increased health complications. This study explores the application of machine learning techniques for diabetes prediction, leveraging a dataset comprising key clinical parameters such as blood glucose levels, BMI, and HbA1c, as well as personal information parameters such as age, gender, smoking history, hypertension and heart disease. Two datasets were used to gain a better understanding of the patterns among Bangladeshi diabetic patients. One dataset consists of collected data, while the other is a combination of the collected data and a dataset from Kaggle. Various machine learning models, including Logistic Regression, Decision Trees, Random Forest, Support Vector Machines (SVM), and XGBoost, were evaluated for their predictive accuracy. Experimental results of the combined datasets indicate that ensemble models, particularly Random Forest and XGBoost, achieved the highest accuracy, exceeding 97% precision. The findings highlight the potential of AI-driven predictive analytics in enhancing diagnosis, optimizing resource allocation, and supporting data-driven decision-making in healthcare. Future advancements in this field may integrate only Bangladeshi diabetic patients’ dataset from wearable devices and electronic medical records, paving the way for multi-disease prediction and improved patient outcomes
E-commerce and its Benefits
(Comilla University, 1-Feb-2025) Hossainv, MD Mozammel
ENHANCED NETWORK SECURITY SYSTEM USING FIRWALLS
(Comilla University, 1-Feb-2025) Sadik, Abdulla Al
The Internet and computer networks are exposed to an adding number
Enhancing Cybersecurity with Biometric Systems and Blockchain Integration
(Comilla University, 2-Feb-2025) Akter, Salma
Strong and safe identity management systems are essential in the age of digital transformation to