MPhil Thesis

Browse

Now showing 1 - 20 of 418

16th Amendment is the Violation of Basic Philosophy of the Constitution: A Critical Analysis
(Daffodil International University, 2018-12-11) Moni, Ranu
The Constitution is the guide which leads a nation to the prosperity. A modern state cannot be thought of without it. So, after nine months long blood-shedding battle in 1971, Bangladesh achieved her long awaited independence and therefore, took an instant effort to formulate a constitution rapidly, based on the ideological spirit of the war of independence. However, to accommodate the demands and will of the people and even sometimes to fulfill the narrow interests of the rulers, Bangladesh Constitution has been amended several times. 16th amendment is one of them. Except a few cases almost every amendments has a great political impacts in the constitution. The most aspired and comprehensive sixteenth amendment induced a great debate among the political parties, intellectual part, constitution experts and masses. This research on An Empirical Study on 16th amendment regarding the impeachment of judges. My research is divided into seven chapters. Our research Paper discussion clears the validity of the 16 amendment, and the arguments of parties. The 16thamendment enacted on 17th September 2014, vested the power of impeachment of judge to the parliament. A writ petition was filed against this amendment and the high court division declared the amendment is void, illegal and unconstitutional. Analyzing the 16th amendment and verdict on 16th amendment it is said that the amendment injured the separation of power, independence of judiciary, rule of law, which are the basic philosophy of the Bangladesh constitution.
A Brief Overview on Electrical Power Generation in Shahjalal Fertilizer Company Limited and Overhauling in the Power Generating Unit
(Daffodil International University, 2019) Arefin, Shamsul
The production activities of several important Industrial sectors in Bangladesh depend on an inadequate supply of natural gas. Natural gas is the main raw material for urea manufacture and basic components for production of urea, ammonia and carbon dioxide. First of all natural gas is cracked by steam in the primary and secondary reformer to produce carbon dioxide, carbon monoxide and hydrogen. Moreover air is also added to the secondary reformer as a source of nitrogen. Carbon monoxide is converted to Carbon dioxide in the shift converters and carbon dioxide is separated from the gas steam in the carbon dioxide removal plant and send into urea plant under high pressure. Remaining components of the gas steam i.e. nitrogen and hydrogen react together under pressure and temperature to from ammonia. This ammonia and carbon dioxide then react together in pool reactor under controlled pressure and temperature to produce urea. Later on this liquid urea is supplied to granulator where spray of liquid urea being bonded with the small amount of binder. The granular urea is bagged with 50 kg content in polythene inserted polythene bags and delivered to the distribution barge and truck and rail wagons from the factory premises.
A Case Study of a Coffee Shop in Dhaka City
(Daffodil International University, 2018-12) Islam, Quazi Khadeemul
The thesis paper has been made on the attempt to use the information of a coffee shop to use them as indicators to the growth and stability of the coffee shop especially of ‘Chayer Cup and Juice House’. The research focuses on the effective practice of financial aspect. This research is limited to Chayer Cup and Juice House’, because the overall objective of this research is to examine the Sales, Purchase, Inventory of ‘Chayer Cup and Juice House’.
A Case Study on Career Decision Making Difficulties and a Prototype Implementation of a Solution
(Daffodil International University, 2018-12) Chowdhury, Asive
Career Decision Making is essential for the human beings to become a successful human being. Most of the people are thinking about this kind of planning since their childhood. In Bangladesh, there are needed to analyze significant features of career decision making like many other countries around the world. The goal of this work to investigate significant features which are responsible to increase decision difficulties factors considering the family income of adolescents. In this situation, we gathered adolescent records of high school going adolescents at Faridganj, Chandpur, Bangladesh using Career Decision Difficulties Questionnaire (CDDQ) which is proposed by I. Gati et. al. In addition, our primary dataset was splitting into 18 datasets based on 3 major CDDQ categories and 6 individual family income ranges. Thus, 10 regression algorithms were applied in our primary dataset and find out the best regression algorithm from them. After that, our selected best algorithm was implemented throughout 18 different datasets and interpreted these findings. In this experiment, we finally observed that adolescents of middle-class families are faced more career decision difficulties than high-and low-class families. We suggest this analysis as a complementary tool for further psychological treatment about career decisions.
A Cloud Based Data Integrity and Confidentiality System
(Daffodil International University, 2019-12-19) Chowdhury, Showmik Zaman
Distributed computing alludes to a framework wherein data preparation and capacity can appear in some good ways from any gadget. Research appraises that supporters worldwide will arrive at 15 billion through the finish of 2014 and 18 billion by means of at the completion of 2016. Because of the developing utilization of gadgets the necessity of distributed computing in gadgets ascend, which advances Cloud Computing. All contraptions require tremendous carport usefulness and most extreme CPU speed. As we're putting away records on the cloud there can be an issue of certainties security. As there might be a chance identified with information carport numerous IT specialists are not indicating their enthusiasm for the heading of Cloud-Computing. To guarantee the clients' records rightness inside the cloud, appropriate here we're giving a powerful component striking element of data uprightness and privacy. This methodology proposed an answer which utilizes the AES set of rules and component of hash work together with different cryptography contraption to offer better security to the data put away on the cloud. This model can't best settle the issue of carport of monstrous realities, anyway also guarantee that it will give information get passage to oversee instruments and ensure sharing records documents with secrecy and uprightness.
A CNN-Based Melanoma Skin Cancer Detection and Classification Approach
(Daffodil International University, 23-01-18) Akter, Farzana
Among various classes of skin cancer, Melanoma is a perilous pattern of skin cancer. Melanoma, widely familiar as malignant melanoma begins in cells which are called melanocytes. From ancient times, people are affected more by that right now. To overcome the complementary problem easier, need to accomplish Melanoma detection as soon as earlier. According to the keen observance and larger analysis of melanoma, CNN achieves better performance both for detection and classification efficiently, specifically deep learning feature-based Convolutional Neural Network, which has the automatic proficiency of skin cancer detection. The proposed method classifies melanoma into two classes, namely Malignant and Benign Melanoma, based on multitasking python libraries. In this research work, the process is come to an end by using the novel CNN model, working with both the training dataset at first and the testing dataset later which has been taken from kaggle platform and is publically available for 10000 images. In the report of the skin cancer image dataset, the experimental results demonstrate a higher level of accuracy rate from the image classifier.
A Collaborative Platform to Collect Data for Developing Machine Translation System
(Daffodil International University, 2018-12-11) Hasan, Md. Arid
The emergence of neural machine translation techniques has opened up a new era for developing translation systems. However, it requires a very large amount of parallel corpus, which is scarce for many under-resourced languages, e.g., Bangla. In order to develop a corpus, currently, there is a lack of publicly available collaborative system. In this paper, we report an online collaborative system for the development of the parallel corpus. The system is developed for supporting any language, however, we only evaluated for developing Bangla-English parallel corpus. In a task completion evaluation experiment, the system outperforms the widely used offline system i.e., OmegaT.
A Comparative Analysis of Credit Card Fraud Detection Using Machine Learning Classification Algorithm
(Daffodil International University, 2021-02) Hossain, Md. Sazzad
The focus of this study is on the design, implementation and management of an ISP network. An Internet Service Provider Business ' main aim is to guarantee a minimum downtime and full network service reliability. ISPs can offer a range of services, including data interconnection, internet service, surveillance, fiber optic networks, network design and repair, and assistance Internet service provider (Internet service providers). Even when there is an unplanned interruption, ISPs have redundancy communication in the company network networks. If power failure or load shedding takes place, the full effectiveness of a built-in network is difficult to ensure. If we can create a network that is reliable, secure, reliance able and quicker, it will be of more use to any organization. Thus, by using robust networks, using renewable energy sources and carefully monitoring we will guarantee optimum network capacity, the network will also improve the quality of service.
A Comparative Analysis of Deep Learning Approaches for Fish Disease Identification
(Daffodil International University, 23-01-18) Sadia, Mobassera Asma
Bangladesh's fisheries and aquaculture industries play a significant role in the nation's food production, ensuring the food supply's nutritional security, growing agricultural exports, and employing 17 million people across various occupations. Farmers who farm fish face a lot of economic losses every year because of various diseases that can happen to fish. There are three common diseases of fish. They are known as black spots, red spots, and white spots. A parasite causes black spot disease. Red spot disease is also known as Epizootic ulcerative syndrome (EUS). And it is caused by a fungus. The white spot syndrome virus causes white spot disease. If a fish farmer can detect these diseases early and apply appropriate treatment, it may protect much infected fish and prevent economic loss. The manual approach of human visualization is a laborious effort for detecting and monitoring fish disease. As a result, any viable strategy that is quick, accurate, and highly automated encourages interest in this problem. Due to a lack of information and a high level of competence, there hasn't been a single piece of useful research on the fish disease. Our system provides solutions to this problem. Fish disease identification using deep learning from images is an arduous task. This study proposes a multi-classification deep learning model for identifying fish diseases (black spots, white spots, red spots, and healthy) from images. To classify and identify these diseases, we will apply five different pre-trained models (DenseNet121, MobileNetV2, ResNet101V2, ResNet152V2, and VGG16), and we have also compared their accuracy. According to the experiment data, the MobileNetV2 model performed better than the other proposed models. In comparison to other models, the model provided good detection accuracy. Keywords— Fish disease identification, Classification, Tensor Flow, Dataset, Deep learning.
A Comparative Study of Different Segmentation Algorithms on Local Fruit Images
(Daffodil International University, 2018-12) Mondal, Diponkar; Supriya, Soummo; Rahman, Mumtahina
This project is on “A Comparative Study of Different Segmentation Algorithms on Local Fruit Images”. This paper represents the different types of segmentation algorithms on various local fruit images. In our whole work we will do various types of segmentation and the segmentation proceed on different types of local fruits shape, the lights flickered and changes color of different local fruits and the noticeable changes when the images will be segmented. The whole system contains four gradations: captured image, preliminary processing, segmentation, and Performance matrix calculation. In captured series we captured various types of local fruits with the device like as camera, in preliminary processing we diminish the noise and after that filter the noisy image to get appropriate image, in segmentation series we will work on four different types of procedure such as K-means clustering algorithms classification, Histogram base segmentation, Otsu method and Thresholding. This all segmentations are our prime need to make this paper. Arrangement of tests were carried out utilizing the proposed demonstrate on a dataset of 5 natural local fruits. Comes about of carrying out these tests illustrate that the proposed approach is competent of consequently recognize the natural product title with a tall degree of exactness.
A Comparative Study Of Lung Cancer Detection Using Deep Transfer Learning With Keras Tuner
(Daffodil International University, 2024-02-04) Datta, Sajeeb
Lung cancer is a type of cancer that begins in the cells of the lungs. It is one of the most common forms of cancer worldwide and is a leading cause of cancer-related deaths. Lung cancer usually develops in the cells lining the air passages of the lungs.There are two main types of lung cancer: non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). NSCLC is the most prevalent, comprising about 85% of all lung cancer cases, while SCLC is generally more aggressive and tends to spread quickly. The need for early detection is underscored by the fact that lung cancer symptoms often manifest at advanced stages, limiting treatment options and reducing the likelihood of successful intervention. This thesis presents a comprehensive study on the application of six pre-trained convolutional neural network models, namely MobileNetV2, InceptionV3, ResNet50, VGG16, VGG19, and NASHNetMobile, for the classification of lung cancer categories. The dataset used in this research consists of 15,000 images, spanning three distinct classes: Lung Adenocarcinoma, Lung Benign Tissue, and Lung Squamous Cell Carcinoma. To optimize model performance, hyperparameter tuning is employed using the Keras Tuner framework. This approach allows for the systematic exploration of hyperparameter configurations to enhance the models' accuracy and generalization. The hyperparameters include learning rates, dropout rates, and other key parameters crucial for model training. The results indicate that MobileNetV2 achieved the highest accuracy among the tested models, with an impressive 98.47%. Following closely, VGG16 demonstrated the second-best performance, achieving an accuracy of 98.40%. The study contributes valuable insights into the practical application of deep learning models for medical image classification tasks, particularly in the context of lung cancer diagnosis. The reported accuracies demonstrate the potential of leveraging pre-trained models to enhance the efficiency and accuracy of computer-aided diagnostic systems for early detection of lung cancer.
A Computer Vision Method for Detecting Traffic Accidents
(DAFFODIL INTERNATIONAL UNIVERSITY, 2024-07-14) Bonny, Afrin Jaman
The study of computer vision makes it possible for machines to replicate the visual system of humans. It is a type of artificial intelligence that gathers data from movies or digital photos and manipulates them to specify the properties. Throughout the process, images are acquired, screened, analyzed, and identified as the information is extracted. Computers can recognize and respond appropriately to visual input thanks to this sophisticated processing. This study's major objective is to utilize computer vision to identify traffic accidents, as these incidents frequently result in fatalities. One of Bangladesh's key problems is the country's high rate of traffic accidents, which is a worldwide problem that impacts more than just Bangladesh. The majority of traffic accidents happen abruptly and without warning, having a major negative impact on both human activity and traffic flow. It is vital to identify traffic incidents as soon as possible and alert oncoming motorists since in the majority of instances, subsequent accidents may be prevented if only prompt identification and prompt rescue were permitted. The organizations' lack of coordination and the careless driving that goes along with it are the causes of these circumstances. Sadly, the primary cause of the high fatality rate in road accidents is the delayed delivery of emergency assistance to accident victims. An accident sufferer may be left neglected for an extended period on highways with light and quick traffic, which might lead to the deaths of those involved in the accident. This paper employs computer vision in detecting accidents from image frames of CCTV footage. The primary goal of this work is to categorize video frames into accident and non-accident categories by training a deep-learning convolution neural network model using each frame of the movie. Convolutional Neural Networks are a quick and accurate method of categorizing photos; with comparably smaller datasets, CNN-based image classifiers have demonstrated an accuracy of over 89%.also discusses the challenges, risks and future uses of ChatGPT and Gemini for ensuring applicable outcomes in education sector.
A Conceptual Framework for Enhancing ATM Security through Integrated Multilayer Biometric Authentication
(DAFFODIL INTERNATIONAL UNIVERSITY, 2024-08-24) Ronok, Koushik Vadury
In recent years, the security of Automated Teller Machines (ATMs) has been a critical concern due to increasing fraud and unauthorized access incidents. This research paper presents a novel multi-layered security framework for cardless ATMs that leverages facial recognition, fingerprint scanning, and One-Time Password (OTP) verification to enhance transaction security. The proposed system aims to eliminate the vulnerabilities associated with traditional card and PIN-based authentication methods. The system operates in three stages: first, the customer's face is scanned and matched against a database to retrieve account details; second, the customer's fingerprint is scanned for authorization; and third, an OTP is sent to the customer's registered mobile device for final verification before the transaction can be completed. This multi-layered approach ensures robust security by requiring multiple forms of identification, making unauthorized access exceedingly tricky.The proposed system provides several benefits, such as improved security via biometric authentication, greater user convenience by removing the necessity for physical cards, and real-time authorization using OTP verification. Nevertheless, implementing this system also comes with challenges, including substantial initial expenses, potential privacy issues concerning biometric data, and users' need to adapt to new authentication methods.This paper delves into developing and deploying a multi-layered security system, assesses its effectiveness through simulation tests, and contrasts it with existing security protocols. The findings reveal that the proposed system significantly enhances ATM security, providing a practical solution to combat fraud and prevent unauthorized access in contemporary banking.ssThis paper examines creating and deploying a multi-layered security system, assesses its performance through simulations, and contrasts it with current security solutions. The results indicate that the proposed system markedly enhances ATM security, presenting a feasible method to guard against fraud and unauthorized access in today's banking industry.
A CRITICAL ANALYSIS OF VIOLANCE AGAINST WOMEN AND GIRLS IN THE DIGITAL AGE OF BANGLADESH
(Daffodil International University, 2019-04-18) ISLAM, MD.JAHIDUL; HRIDAY, BAPPA BARUA
Research is essential for pursuit of higher study and acquiring scientific knowledge in any discipline. In-depth study of law is hardly possible without undertaking research. This Research Dissertation entitled. “ADDRESSEING VIOLANCE AGAINST WOMEN AND GIRLS IN THE DIGITAL AGE OF BANGLADESH: A CRITICAL ANALYSIS” Has been done by us as a part of the LLM (masters) course curriculum. This research is intended to a comprehensive analysis of the existing situation in our country. The basic purpose of this research is to find out the current functions undertaken by the Government, the legal profession. The intention behind this research is to clarify to the Government for enact laws for the protection and security of our women who are working in different sector for their livelihood, studying in different places, using internet for better communication and gathering knowledge by using internet. We do not claim that our views and observations are correct from every point of view. There may be shortcomings and wrongs which all are mine and we are responsible for those. We should also acknowledge that we alone had composed the whole research dissertation with very short time. So we hope that readers will take into consideration all these and bear with errors and other shortcomings for better research dissertation. Any Enquiry on the contents of this Research dissertation will be most welcome by us.
A Critical Overhaul of the Laws and Their Trend Regulating the Cyber-crime in Bangladesh
(Daffodil International University, 2019-09-14) Showrov, Mohsin Uddin
We know that the present age is the technological age. The world is highly changing day by day. So, the modes of crime are also changing now. In Bangladesh, Cyber and technology related crimes are gradually increasing. At present, Cyber-crime becomes the most serious issue in Bangladesh. Already it has been seen that cyber-crime is a despondent warning which may become visible in the area of information and communication technology. In addition, we also saw that Cyber blustering is becoming a major anxiety for parents because of their children using the internet as their part of the subject by which majority of students in Bangladesh have experienced being disturbed online or being disturbed by the same person both their online or offline. Therefore, cyber-crimes are also becoming a warning to the government itself. In our country, there are only few laws to regulating and controlling the cyber-crime but this is not enough to control this type of crimes.
A Decision-Making Approach for Trauma Center Site Selection
(Daffodil International University, 2020-07-09) Sarkar, Amit
Normally when one’s made a trauma center then once select a suitable place. For selection a place we need to follow many procedures. And need some specialized man power. Such as architect, urban engineer, civil engineer, medical technician etc. For those we need many times, coast and man power. That’s all are problem. To solves those problem and saves time & money we make a research project which name is A Decision-Making Approach for Trauma Center Site Selection provides a platform which helps to saves time, money, man power and solve those problems. Trauma Center Site Selection is a selection process which is select a suitable place from five or six places, this process calculated by AHP (Analytical Hierarchy Process) algorithm. By using this algorithm also select mobile, car, motorcycle, flat and any types of selection.
A deep learning approach for classification of liver disease
(Daffodil International University, 2024-01-24) Kalam, Faria Binta
Liver diseases pose a significant global health burden, with diverse manifestations such as ballooning, fibrosis, inflammation, and steatosis. Accurate and timely diagnosis is crucial for effective treatment planning and patient management. This thesis explores the application of deep learning models, including EfficientNetB2, VGG16, InceptionNetV3, DenseNet121, and ResNet50, for the comprehensive classification of liver diseases based on these distinct pathological features. The study involves a robust dataset of liver pathology images, capturing various stages and manifestations of liver diseases. Through an exhaustive analysis, we compare the performance of different deep learning architectures in accurately identifying and classifying ballooning, fibrosis, inflammation, and steatosis. Our experiments reveal that EfficientNetB2 outperforms the other models in terms of accuracy, demonstrating its efficacy in handling the complexities of liver disease classification. In addition to model performance, the thesis delves into interpretability, providing insights into the features and patterns learned by each model. This contributes to a better understanding of the decision-making process and enhances the clinical relevance of the deep learning models in real-world scenarios. The findings of this research not only showcase the potential of deep learning in liver disease diagnosis but also highlight the significance of selecting appropriate architectures for optimal results. The implementation of EfficientNetB2 in this context opens avenues for improved diagnostic tools and automated systems that can aid healthcare professionals in making more informed decisionsfor patients with liver diseases. The implications of thisstudy extend beyond liver disease classification, emphasizing the broader applicability of deep learning in medical imaging and pathology. The insights gained from this research contribute to the ongoing efforts to enhance the accuracy and efficiency of computer-aided diagnostic systems in the field of hepatology.
A deep learning approach for classification of liver disease
(Daffodil International University, 2024-01-25) Kalam, Binte Faria
Liver diseases pose a significant global health burden, with diverse manifestations such as ballooning, fibrosis, inflammation, and steatosis. Accurate and timely diagnosis is crucial for effective treatment planning and patient management. This thesis explores the application of deep learning models, including EfficientNetB2, VGG16, InceptionNetV3, DenseNet121, and ResNet50, for the comprehensive classification of liver diseases based on these distinct pathological features. The study involves a robust dataset of liver pathology images, capturing various stages and manifestations of liver diseases. Through an exhaustive analysis, we compare the performance of different deep learning architectures in accurately identifying and classifying ballooning, fibrosis, inflammation, and steatosis. Our experiments reveal that EfficientNetB2 outperforms the other models in terms of accuracy, demonstrating its efficacy in handling the complexities of liver disease classification. In addition to model performance, the thesis delves into interpretability, providing insights into the features and patterns learned by each model. This contributes to a better understanding of the decision-making process and enhances the clinical relevance of the deep learning models in real-world scenarios. The findings of this research not only showcase the potential of deep learning in liver disease diagnosis but also highlight the significance of selecting appropriate architectures for optimal results. The implementation of EfficientNetB2 in this context opens avenues for improved diagnostic tools and automated systems that can aid healthcare professionals in making more informed decisionsfor patients with liver diseases. The implications of thisstudy extend beyond liver disease classification, emphasizing the broader applicability of deep learning in medical imaging and pathology. The insights gained from this research contribute to the ongoing efforts to enhance the accuracy and efficiency of computer-aided diagnostic systems in the field of hepatology.
A Double Key-Based Public-Private Key Encryption-Decryption Process for Secured Message Transaction
(Daffodil International University, 2019-12-06) Rashiduzzaman, Muhammad
In modern communication age, security of electronic message transaction is the demand of time. It is most essential in various aspects. Currently a large amount of sensitive data is transmitted over the open network or internet or other communication channels on a daily basis. Without strong security, we cannot protect these sensitive information from malicious attacks. Currently, it is main concern to impose additional security services to the communicating message, communication channel and communicating participants. For this, a better approach for electronic message transaction system has been developed using Python programming language. It performs electronic message transactions with all the fundamental security requirements, which are confidentiality, integrity, authentication and non-repudiation for both communicating message and communicating participants. To do this, simple cryptographic encryption and decryption techniques are used to the communicating messages. At first message is encrypts with the private key of sender PRa and the output is again encrypts with a shared secret key K1 that generates cipherext, which is again encrypts with another shared secret key K2 that generates a code that serves as message authenticator known as MAC, which is concatenates with the ciphertext and again encrypts them with shared secret key K1 that builds the new cphertext, which is again encrypts with the receiver’s public key PUb to produce final ciphertext that is to be send to the intendent recipient. In the receiving end, to retrieve the message, receiver at first decrypts the received information with his private key PRb and again decrypts with the shared secret key K1 that gives the ciphertext and MAC of the ciphertext, and then only decrypts the MAC to generate a new ciphertext′ and compare the new ciphertext′ with the received ciphertext that ensures the ciphertext authentication as well as message authentication; if ciphertexts are found same, then decrypts the ciphertext with shared secret key K1 and again decrypts with the sender public key PUa and retrieve the message; otherwise discard it. This technique can be applied anywhere of electronic communications in a secure fashion.
A Framework for Human Skin Disease Classification Using Convolutional Neural Network
(Daffodil International University, 2025-01-20) Hera, Mst. Dilruba Yeasmin
One of the most dangerous types of cancer is skin cancer, it becomes a significant health hazard when not treated and detected on time. Skin cancer may spread to other parts of the body and complicate treatment if it is not detected in its early stages. Mainly it is the result of abnormal skin cell growth, usually the cells are stimulated by the sun for a long time. The early detection of skin tumors is a basic but highly complicated and expensive process due to the complexity of the diagnostic methods implicated. The identification of skin cancer by the location and cells involved augments the necessity of a very precise classifier for a successful diagnosis. Where the use of CNN in the recognition and classification of skin cancer, especially in skin lesion classification has been proposed to solve this issue. The utilized diagnosing method includes the utilization of image processing algorithms and deep learning models to increase accuracy and efficiency. Methods like image augmentation are then used for adding more rows to the dataset are used to scale up the dataset. This way, the model understands the diverse cases encountered. In addition, transfer learning is useful for increasing the classification accuracy by using pre-trained models for improved performance. As one of deep learning's deep architectures, CNNs serves as a key player in the extraction of features and in the classification of skin problems like psoriasis. This technique has been impressively productive for it gets a hit rate of 75%, thus revealing future prospects in the medical field.