http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/5543 Sun, 07 Jun 2026 23:53:59 GMT 2026-06-07T23:53:59Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17229 A Hybrid Machine Learning Model for Enhanced Prediction of Gestational Diabetes Using Diverse Datasets Miti, Ramisa Maliyat Gestational diabetes mellitus (GDM) is a significant health concern affecting maternal and fetal well-being, necessitating early and accurate predictive models. This study presents a novel hybrid machine learning model integrating Random Forest, Support Vector Machine, and Gradient Boosting Machine through a stacking ensemble approach. The hybrid model achieved superior performance across two datasets, with accuracy scores of 92.7% and 89.02%, significantly outperforming individual models. The integration of diverse data sources, including clinical, biochemical, and demographic variables, enhanced the model's robustness and generalizability. Metrics such as precision (91.5% and 86.05%), F1-Score (92.3% and 73.18%), and ROC-AUC (0.94 and 0.91) underscore the model's ability to balance precision and recall effectively. The study addresses key research gaps, including generalizability issues, data integration, and scalability. By incorporating hyperparameter tuning, model pruning, and quantization, the hybrid model is optimized for deployment in resource-constrained settings, demonstrating scalability and efficiency. Despite its promise, challenges such as the need for external validation across diverse populations and addressing biases in training data remain. Future research should focus on fairness-aware algorithms and longitudinal studies to ensure equitable healthcare outcomes. This hybrid model showcases its potential as a reliable tool for early GDM detection, enabling timely interventions and improving maternal and fetal health outcomes. Its integration into clinical workflows and adaptability across healthcare settings highlight its significance as a step forward in precision medicine. Thesis Report Sat, 11 Jan 2025 00:00:00 GMT http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17229 2025-01-11T00:00:00Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17228 nderstanding AI tool engagement: A study of ChatGPT usage and word-of-mouth among university students By PLS Algorithm Toma, Hozaifa Afroz This study examines how university students use ChatGPT, focusing on factors including system quality, personalization, trust, and satisfaction. This study examines the effects of these factors on students' behavioral intentions and word-of-mouth recommendations. Partial Least Square (PLS, is a method for survey and data analysis. The results show how crucial customisation and trust are to increasing user satisfaction and encouraging repeat use. Word-of-mouth is discovered to have a substantial impact on this adoption. Even with ChatGPT's many benefits, Data privacy issues still remain despite ChatGPT's several advantages. It provides suggestions for enhancing AI driven educational resources, notably around the user interface, ensuring it is safe, visualised, relevant and of great quality. Thesis Report Sat, 11 Jan 2025 00:00:00 GMT http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17228 2025-01-11T00:00:00Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17223 Explainable AI-Based Anemia Prediction through Machine Learning Sohan, Md Sohanur Rohaman Anemia is a common worldwide health problem that can have a major negative impact on a person's quality of life and result in major health issues. Timely intervention depends on early diagnosis and precise Anemia prediction however conventional diagnostic techniques frequently depend on intrusive procedures or subjective interpretation. This thesis investigates the use of machine learning approaches to forecast anemia through the application of explainable artificial intelligence (XAI). In particular, it seeks to create an interpretable model that not only forecasts the risk of anemia but also offers intelligible information about the variables influencing the prediction. A variety of machine learning algorithms are used in this work to optimize the accuracy and reliability of anemia prediction. A robust and flexible approach is ensured by the unique benefits that each of the selected methods—Logistic Regression, Random Forest, Decision Trees, XG-Boost (Extreme Gradient Boosting), Support Vector Machine (SVM), and KNN— brings to the prediction process. These models are trained on a clinically relevant dataset that includes hemoglobin levels, red blood cell counts, and other CBC characteristics that are crucial for detecting anemia. This study uses a large dataset of over 1281 people that contains demographic, clinical, and lifestyle factors associated with anemia risk. Carefully chosen machine learning models are trained and evaluated using the preprocessed data. Predefined metrics like F1 Score, accuracy, precision, and recall are used to assess each algorithm's performance. When compared to other algorithms, Decision Tree performs better than all others in terms of prediction accuracy, with a remarkable 99.03% in our analysis. This implies that decision trees are a better way to forecast anemia. Decision trees' outstanding performance enables the creation of a precise anemia prediction tool. By analyzing readily available patient data, this method can assist medical professionals in preventing anemia and initiating therapy early. Thesis Report Tue, 14 Jan 2025 00:00:00 GMT http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17223 2025-01-14T00:00:00Z http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17219 Fine-Tuning Large Language Models For Depression And Anxiety Detection On Twitter Bhuiyan, MD Siyam User mental states are represented via social networking sites like Twitter, which also record the users thoughts and emotions. This transformation extends to fields like public health epidemiology, where analyzing social media has become a valuable tool for understanding mental health trends. Anxiety and depression remain the most widespread mental health issues globally, with their prevalence growing over the last decade.(Organization, ) Social media,especially Twitter,offers a unique glimpse into individuals mental states,and researchers have found that analyzing tweets can provide real-time insights into shifts in mental well-being.(Choudhury, Counts, & Horvitz, )The integration of artificial intelligence, especially natural language processing (NLP),has opened doors to processing massive amounts of unstructured text. These advancements make it feasible to create systems that monitor mental health at scale, offering a practical approach to identifying and addressing mental health concerns through digital footprints. Thesis Report Tue, 21 Jan 2025 00:00:00 GMT http://dspace.daffodilvarsity.edu.bd:8080/handle/123456789/17219 2025-01-21T00:00:00Z