M.Phil / MS
Permanent URI for this collectionhttps://repository.cuilahore.edu.pk/handle/123456789/36
This collection archives the complete set of theses produced by students of the COMSATS University Islamabad, Lahore Campus.
Browse
13 results
Search Results
Item Magnification Independent Approach to Diagnose the Breast Cancer from Histopathological Images(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2022) SHEHROZ TARIQ; SP20-RCS-017; LHR TP 8354; Dr. Allah Bux SarganoBreast cancer is one of the top cancers that cause death globally. Hematoxylin and Eosin-stained images diagnose biopsy tissue, and experts are usually upset with the closing opinion. Computer-assisted diagnosis techniques enable to cut costs and improve efficiency in this practice. Automation of breast cancer multi class classification from microscopic images has a significant impact on computer-assisted breast cancer identification or prediction. The purpose of breast cancer multiclass classification is to classify different subtypes of breast cancer (Papillary, Adenosis, Mucinous, etc.) after identifying the Benign or Malignant Class. Yet, multi-class classification of breast cancer from microscopic imageries aspects dual major encounters: First, the excessive problems in methods of breast cancer multi-class classification compared to binary class classification (benign vs. malignant), and second, the slight variances in several classes because of the high-quality image inconsistency forms, high cohesiveness of malignant cells, and inconsistency of color distribution. As a result, although the automatic multi-class classification of breast cancer from microscopic images has considerable clinical importance, it has never been investigated. As we are dealing with microscopic imaging, magnification plays a vital role while classifying cancer. Existing literature techniques exclusively emphasize magnification-dependent binary or multi class classification and do not continue the effort for magnification-independent breast cancer diagnosis. Using a proposed model inspired by Vision Transformer (ViT) model with multiple variations, this work offers a robust and novel breast cancer multi-class classification approach combined with a binary classification method from a clinical standpoint. The evaluation methods that are utilized in our work are the accuracy score and confusion matrix. The comprehensive experiments are conducted using the publicly accessible benchmark breast cancer dataset named as BreaKHis. As we have trained three models; first for binary classification and the other two for multiclass classification. We achieved 89.4% accuracy for our first model, 74.57%, and x 57.41% accuracies for the remaining two multi-class models. These outcomes are astonishing as related to existing literature. As existing literature technique claims 88.9%, 63.6%, and 52.7% results against these models. The outcomes showed that the proposed methodology surpassed existing methods in breast cancer diagnosis by a significant margin. This is because of the Convolutional layers, we added before passing to the transformer in our proposed model. This convolution allows the features of the breast legion to be further prominent before applying the multi-head attentionItem Identifying Multilingual Textual Cyberbullying on Social Networking Using a Novel Machine Learning Approach(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2022) Manahil Fatima; SP20-RCS-022; Dr Adnan Ahmad; LHR TP 8347Technological advances have made the internet accessible to the vast majority of people. In the current era, the usage of social networking sites or media and discussion forums as a form of communication has increased dramatically. Trolls that spew bully, hate, and toxic speech on social media and in forums will also be able to use this new technology. The identification of cyberbullying on the internet is becoming an increasingly important topic. During the last decade, Poisonous comments have been sarcastic, rude, insulting, and bullying, resulting in 26% of people thinking about suicide or harming themselves, and many go through depression and anxiety. Methods currently used to address Internet toxicity are difficult to measure since it relies so heavily on human moderators larger than necessary to handle a growing user base. The author researched the detection of cyberbullying, which has become a significant part of the internet and affects internet users around the globe; this research is done in three different languages. These languages include English as International Language, Spanish as the 3rd most used online language, and Roman Urdu as the national language used in India to communicate on the internet. With this in mind, the author will begin by employing a wide range of established classical machine learning and cutting-edge techniques and Incorporating techniques from deep learning into four different datasets with a distinct sizes. The research will be conducted as part of this project's goal to start; there is a need to develop an ensemble model that can be used to identify the hazardous language. Two different ensemble models are proposed; one from machine learning classifiers includes Support Vector Machine, Naive Bayes, Random Forest, and Logistic Regression using Feature extraction. The second ensemble learning model is on deep learning techniques using hyperparameter tuning of CNN, LSTM, and GRU. It also compares their results with Transfer learning classifiers, including BERT and Distil BERT. Distil BERT gives 100% accuracy, recall F1-score, and precision on all four different datasets. In English x Dataset 1 Unigram feature gives 74% accuracy, and 3-3 char gram provides 77% accuracy. In CNN, using Adam optimizer with 32 batch size and 8 epochs at 0.01 learn rate gives the highest accuracy of 56%, while in LSTM, rmsprop optimizer gives the highest accuracy of 53% with batch size 32 and 8 epochs at 0.001 learn rate. In GRU, the rmsprop optimizer gives the highest accuracy of 52% with batch sizes 32 and 16 epochs at a 0.001 learn rate. The ensemble model of machine learning provides the highest accuracy of 65%, while the deep learning ensemble model gives 57% accuracy. In English Dataset 2 Unigram feature gives 95% accuracy, and 4-4 char gram provides 95% accuracy by random forest. In CNN, using the rmsprop optimizer with 32 batch sizes and 16 epochs at 0.001 learn rate gives the highest accuracy of 94%, while in LSTM, the rmsprop optimizer gives the highest accuracy of 92% with batch size 64 and 16 epochs at 0.01 learn rate. In GRU, the rmsprop optimizer gives the highest accuracy of 91% with batch sizes 64 and 16 epochs at a 0.01 learn rate. The ensemble model of machine learning provides the highest accuracy of 87%, while the deep learning ensemble model gives 80% accuracy. In the Spanish dataset, the Unigram feature gives 78% accuracy, and 5-5 char gram provides 80% accuracy by Logistic Regression. The highest 76% accuracies are gained at batch size 64, epochs 8 learn rate 0.001, and optimizer Adam in CNN, while 75% using batch size 64, epochs 8, learn rate 0.001, optimizer rmsprop in LSTM. In GRU, 74 % accuracy is gained using batch size 64, 'epochs 8, learn rate 0.01, 'optimizer rmsprop. The ensemble model of machine learning provides the highest accuracy of 79%, while the deep learning ensemble model gives 74% accuracy. In the Roman Urdu dataset, the Unigram feature gives 80% accuracy by Naïve Bayes, and 5-5 char gram provides 80% accuracy by Random Forest. The highest 76% accuracies are gained at batch size 32, epochs 16, learn rate 0.001, and optimizer Adam in CNN, while 75% using batch size 32, epochs 8, learning rate 0.01, optimizer rmsprop in LSTM. In GRU, 74 % accuracy is gained using batch size 64, 'epochs 8, learn rate 0.01, 'optimizer rmsprop. The ensemble model of machine learning provides the highest accuracy of 79%, while the deep learning ensemble model gives 74% accuracy.Item Repeated Buyer Prediction: A Study of Repurchasing Intention of Buyer in E Commerce(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2022) Muhammad Usman; SP20-RCS-014; Dr. Hamid Turab Mirza; LHR TP 8358During promotions, retailers frequently obtain a huge number of new customers. However, several of the purchasers attracted are one-time hunters, and the deals are unlikely to have a long-term influence on sales. It was critical for merchants to discover who may be turned into regular loyal consumers and afterward target them to reduce promotion costs and boost return on investment (ROI). It was critical for merchants to discover who could be converted into repeat customers to solve this problem. Merchants may significantly cut promotion costs and increase the return on that investment by focusing on these prospective loyal consumers (ROI). Consumer targeting in the area of internet advertising was generally known to be difficult, especially for first-time consumers. In this work, collect a collection of merchants as well as their associated new buyers gained during the "Double 11" day offer using Tmall.com's long-term user behavior record. This experiment objective was to predict whether new clients would become loyal consumers in the future for certain merchants. In other words, this experiment must estimate the probability, which these new purchasers will buying within the similar merchants again for the next six months. This work suggested employing enhanced merged models (XGBoost as well as LightGBM and Histogram-based gradient boosting machine to forecast a repeat customer and feature engineering through extracting feasible features by which important components would be derived to train the model to prophesy the repeated buyer. These experimental findings suggest that when compared to the original models, this work-combined model may achieve significant performance increases. XGBoost accuracy was 95.97 and AUC was 0.9753. The accuracy of Light GBM was 92.58% and the AUC was 96.15%. Histogram-based gradient boosting machine training accuracy was 96.7% and the testing accuracy was 96.27%.Item Analysis and Evaluation of Risk Prediction using Software Requirements(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2022) Shanzeh Shaukat; SP20-RCS-011; LHR TP 8335; Dr. Muhammad SharjeelSoftware requirements play a pivotal role in the planning of a software project and they are a key input to project size and effort estimation processes. The increasing dependence of all aspects of life on software products has made the success of software projects more critical. It has become a challenge to predict software project risks as early as possible and create contingency plans to overcome or mitigate future risks. Furthermore, the risk is different for different projects. Currently, there is no consensus within the software development community on requirement properties to determine the cause of risk in software projects. There is a need for a systematic analysis that combines risk prediction factors, risk prediction methods, risk datasets, and their analysis in a single framework. This research aims to identify requirement risk attributes and project success factors that are later on validated using SEM (Structural Equation Modeling) to analyze that requirement risk properties can be linked to software risks. Mainly, the purpose of conducting a systematic mapping study (SMS) is to find out the properties of requirements that are the major cause of risk occurring in software projects. Later, we designed a survey based on the requirement properties extracted from SMS to find out risk prediction practices used by software professionals and their views related to these factors in the Pakistani software industry. The proposed SMS and survey in the Pakistani industry based on the risk prediction process using SEM is missing in the published research. That’s why, this research will help project managers to understand, initiate and evaluate the risk prediction process in their software development organizations. Additionally, it will help us to identify the gap between research on risk prediction and actual practices in the software industryItem Monocular 3D Object Detection for Autonomous Vehicle(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Wasif Maqsood; SP20-RCS-012; LHR TP 8669; Dr. Aksam IftikharCurrently, Environment perception, 3D objects detection and the distance of objects from the camera is one of the hot topics in computer vision and in robotics, which is widely explored by scientists to achieve maximum accuracy of detection for autonomous vehicles. For reliable and safe driving, it is necessary that self-driving cars can perceive the environmental surroundings accurately. 3D object detection and their distance estimation are a challenging task because of different angles of moving vehicles and computational resources required to process video data. Distance estimation from the camera is used in all autonomous vehicles and robots for safe driving. In this research, a two-stage deep learning architecture is proposed for 3D object detection, their pose estimation and then the distance of objects using monocular cameras installed in vehicles. In contrast to stoneworker methods which only regress 3D dimensions, we propose a method in which using deep neural network we regress 2D bounding boxes, geometric estimation and the distance from the camera and then use these estimations for regressing accurate 3D object properties and estimate pose to construct the stable 3D bounding box. Our models is tested on the KITTI Dataset, which consists of images of vehicles in different environments. The dataset contains separate repositories for training and testing purposes (7481 and 7518 images, respectively) with main target classes (cars, pedestrians).In this Thesis we discussed deep learning techniques for computer vision. More precisely, we are focusing on the 3D bounding boxes and distance estimation from the scene by using only single for autonomous vehicles and robots. In this chapter we present an introductory approach for the problem and also present our contributions and objectives of this thesis.Item Recommendations on Maintainability Issues in Web Services(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Zoha Ejaz; SP20-RCS-031; LHR TP 8671; Dr. Ghulam RasoolWeb services are components that are capable of performing a certain task, like two devices communicating with each other and exchanging information. With the increase in use of web service software, following the good practices for maintainability plays an important role as it helps to reduce maintenance cost and improve software quality. Many problems are faced by the developers nowadays and they need a platform to discuss these problems and get information on how to resolve them by an expert opinion. They ask questions on the discussion forums or the Community Question Answering (CQA) sites such as Stack Overflow and Quora. These problems include: outdated online code snippets, maintainability issues in reuse of existing libraries, RESTful webservices composition. The Representational State Transfer (REST) is an established architectural paradigm utilised in the development and design of web services. The system employs the Hypertext Transfer Protocol (HTTP) to facilitate the transmission and reception of data. It is one of the most used domains and asked about on CQA sites. The state-of-the-art tools and techniques are limited as they mostly cover theoretical knowledge to state RESTful web services patterns. We need to consider the industrial opinion as their opinions can be used to overcome many problems and identify new ones for resolution. A new approach is proposed in which the existing practices defined in the literature are used as a benchmark and expertise opinions are extracted from Stack Overflow accepted answers. Recommendations are formed on the basis of the experts’ opinion and used in a tool which detects the area of discussion and suggests expert’s pattern definitions.Item A Machine Learning Technique for Motion Planning in Articulated Robots(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Moman Ali Haider; SP20-RCS-024; LHR TP 8670; Dr. Wajahat Mahmood QaziThe integration of human-like motion into robotic systems has emerged as a pivotal research frontier, driven by the aspiration to create robots capable of nuanced interactions in real-world scenarios. This thesis delves into the development and implementation of a novel solution, leveraging state-of-the-art motion planning algorithms and neural network based technologies to instill human-like qualities in robotic motion. The research unfolds against the backdrop of contemporary challenges in the field of robotics, particularly the need for motion planning algorithms (MPAs) that not only navigate collision-free paths but also emulate human-like movements. In the landscape of motion planning, traditional algorithms like rapidly exploring random trees (RRT) and its variants have proven effective, yet their computational complexity becomes a bottleneck in higher-dimensional problem spaces. To address this limitation, the motion planning networks (MPNet) paradigm is introduced, utilizing a neural network approach with point cloud representations to navigate the intricacies of higher-dimensional environments. The computational efficiency of MPNet is harnessed to overcome challenges associated with dimensionality, offering a promising avenue for generating collision-free paths. The central problem addressed by this research is the lack of human-likeness in the paths generated by MPAs, limiting their applicability in tasks that demand human-like motion. Prior attempts to imbue human-likeness often relied on datasets recorded from human movements, leading to unreliable and constrained solutions. In response, the proposed solution adopts a hybrid approach, combining the strengths of MPNet and Artificial x VFRRT. The latter is chosen for its ability to generate human-like paths, albeit with limitations in higher-dimensional problems. The research methodology unfolds in distinct phases, beginning with the development of a reliable human-like motion dataset. The Data Generator (DG) module orchestrates this process, employing an Artificial VFRRT-based motion planner within the Kautham simulation tool. The dataset, characterized by dynamic path generation strategies and environmental diversity enhancements, forms the foundational building block for subsequent modules. The Data Encoder (DE) module steps in to transform raw obstacle representations into a point cloud format compatible with MPNet training. This adaptive encoding ensures usability and efficiency, setting the stage for the Human-like MPNet (HLMPNet) module. HLMPNet marks a paradigm shift in motion planning architectures, dynamically adapting learning parameters through an iterative process informed by human-likeness evaluations. This module is designed not only to replicate human motions but to refine and adapt its behavior based on nuanced feedback. The Human-Likeness Evaluator module acts as the discerning judge in the evaluation framework, quantifying the authenticity of generated paths. Its role in continuous learning and optimization ensures that HLMPNet evolves towards increasingly authentic and nuanced human-like motion planning. The significance of this research lies in its practical applications across various domains, including real-world human-robot collaboration, user-friendly interfaces, efficient and safe robotic operations, enhanced experiences in entertainment and services, and improved assistive and rehabilitation technologies. The proposed solution offers a holistic approach to addressing the challenges of human-like motion in robotics, contributing to the ongoing evolution of robotic systems in diverse applicationsItem Investigating Cyclic Translation for Urdu Text Reuse Detection using Deep and Transfer Learning Methods(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Muhammad Arslan; SP20-RCS-025; LHR TP 8466; Dr. Muhammad SharjeelThis research study addresses the complex problem of generating text reuse corpora using Machine Translation (MT), effectively detecting text reuse in paraphrased text through a cyclic translation approach and investigating the behavior of newly proposed deep learning and transfer learning methods on various Urdu text reuse corpora. Prior research endeavors addressing the challenge of text reuse paraphrase detection through the application of deep learning methodologies have predominantly focused on the English language. However, a comprehensive investigation into the problem of text reuse detection in the Urdu language, specifically utilizing a cyclic translation approach in the processing of data for training and testing purposes, alongside the utilization of deep learning methods, remains relatively limited. To tackle these challenges, a translated dataset is constructed by employing the Python Google Translate API on the Counter Dataset. Two variations of deep neural networks, namely Siamese Bidirectional LSTM (BiLSTM) and Manhattan LSTM (MaLSTM), are employed to achieve optimal accuracy in the multi-classification task. The training process incorporates two well-known optimizers, RMSProp (Root Mean Square Propagation) and Adam (Adaptive Moment Estimation). The experimental results substantiate the effectiveness of the Bidirectional LSTM, which achieves an accuracy of 77.22%, and the MALSTM, which attains an accuracy of 76.20%. These findings provide valuable insights for researchers and practitioners, paving the way for further advancements in the study of text reuse and its implications in diverse linguistic contextsItem City Scale Multi Camera Vehicle Tracking Using Deep Learning(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2023) Muhammad Sameed Khan; SP20-RCS-013; LHR TP 8465; Dr. Zeeshan GillaniIntelligent city and traffic management depend on city-scale multi-camera vehicle tracking; however, this work has various difficulties. On the different viewing angles, problems arise that include the variation on the large scale, frequent occlusion and appearance variation. In this study, we use the cross-camera tracking technique and multi-camera tracking system that considers aggregation loss. To address the challenges of multi-camera vehicle tracking, the suggested system has four key parts. First, we extract the tracks with the help of a single camera view by identifying the object and multi-object tracking modules. These modules combine their detection capabilities to provide efficient tracking between frames. After obtaining the tracklets, we use a multi camera re-identification module to match them. The tracklets acquired by several cameras are connected by this module using re-identification techniques. We used OsNetX1_0 and retinaNet50 for reidentification. In the final stage, we deal with the isolated trackless and tracking of the synchronize ids that rely on the outcomes of the re-identification. We improve the computational performance with the help of less parameter models and cohesion of the tracking system by removing isolated tracklets and assuring consistent tracking IDs. With practical and effective alternatives, the research advances multi-camera vehicle tracking on the city scale. However, this system shows the significance of the fast-multi-target cross-camera tracking approaches and the loss of aggregation in the study challenges. The AI city challenges' success reveals our system's potency and viability.Item Good or Bad: Design Principles for Resolving Maintainability Issues for Microservices(Library Information Services, COMSATS University Islamabad, Lahore Campus, 2022) Maham Saleem; SP20-RCS-029; LHR TP 8049; Dr. Ghulam RasoolMicroservices Architecture (MSA) style has gained popularity in the software industry as it provides a powerful method for developing applications that consist of several small, maintainable, independently deployable, and manageable services. The loosely coupled architecture style and dynamic nature of MSA cause many issues such as maintainability, security, latency, and performance. Anti-patterns are "poor" solutions to persistent issues. These solutions frequently make software systems more difficult to maintain. Microservice-based systems experience difficulties with maintainability and evolution, much like any other architectural style, because of these anti-patterns. As a result, adding and modifying functionalities becomes more and more difficult. The maintenance process also includes identifying anti-patterns within microservice-based systems, which improves the software's quality assessment. In this research, we analyze and understand the maintainability issues, anti-patterns, and suggested solutions related to microservices being discussed in the literature, and the Industry Forum. With the help of these blogs and academic literature, we propose a methodology to address the maintainability, design, and architecture issues for the improvement of microservices. We want to provide an extensive list of microservices anti-patterns and tool support to provide suggestions and categorize the best practices proposed by the professionals to support the research community and industry. In conclusion, we seek to improve the maintenance and quality of microservice-based systems with our semi-automatic tool-based approach. Our strategy is based on software re-engineering methods to extract useful data and has been created as an automated approach integrated with an add-in for Sparx System Enterprise Architecture (EA) for anti-patterns recognition from MS-based software applications. We have evaluated our approach on 20 open source microservices-based systems using precision and recall metric