Browse
Recent Submissions
Item Leveraging Simplified Machine Learning for KSE-100 Index Prediction(UMT, Lahore, 2025) Hassaan AhmadThe Karachi Stock Exchange 100 (KSE-100) Index, a benchmark of Pakistan's stock market, reflects the performance of the top 100 companies and serves as a vital indicator of the country's economic health. However, the stochastic and volatile nature of stock markets, coupled with the complexity of analyzing numerous macroeconomic and sectoral variables, makes forecasting the KSE-100 Index a daunting task. Traditional time series models like ARIMA and ARMA, though foundational, struggle to capture the nonlinear dynamics and high volatility inherent in financial data. This research explores the transformative potential of machine learning (ML) techniques in addressing these challenges. By employing Random Forest, Gradient Boosting, and Support Vector Regression, the study evaluates their efficacy in predicting the KSE-100 Index. Using historical data spanning December 2019 to June 2024, the research integrates key macroeconomic variables such as GDP growth, inflation, and trading volume, enhancing the predictive power of these models. Among the models tested, Gradient Boosting emerged as the most effective, achieving an R² of 0.91, highlighting its ability to capture complex patterns and dependencies. Key contributions of this thesis include simplifying predictive frameworks, enabling accessibility for non-experts, and proposing a scalable approach that balances high accuracy with usability. The integration of feature engineering techniques, such as moving averages and volatility metrics, further underscores the robustness of the proposed methodology. Additionally, the research highlights critical gaps, such as the lack of real-time data integration and the need for user-friendly forecasting tools, paving the way for future advancements. This thesis bridges the gap between advanced financial forecasting techniques and practical applications, offering a democratized approach to stock market prediction. By empowering individual investors and small-scale firms, this study contributes to a more inclusive financial ecosystem and encourages informed decision-making in a time of economic uncertainty.Item AUGMENTING PREDICTIVE ACCURACY THROUGH HYBRID INTELLIGENCE: A COMPARATIVE ANALYSIS OF ENSEMBLE LEARNING TECHNIQUES(UMT, Lahore, 2025) MUHAMMAD ZAIN ASHRAFA revolutionary strategy, hybrid intelligence improves prediction accuracy in a variety of fields by fusing artificial intelligence and human experience. This synthesis makes use of AI systems' computing efficiency, scalability, and pattern recognition skills in addition to human cognition's intuitive reasoning, contextual awareness, and ethical foundation. When used with ensemble learning frameworks, hybrid intelligence is very powerful since it allows for the creation of strong, multi-layered predictions by utilizing a variety of data streams from many fields. In order to support and enhance decision-making processes, this article investigates the potential of hybrid intelligence in combining cross-domain data from social media, healthcare, finance, and environmental systems. One of the primary use cases is the stock market, where significant volatility and the impact of numerous international factors have historically made predicted accuracy difficult. Hybrid ensemble models, which combine machine learning approaches, provide layered inference, with human domain experts evaluating, modifying, and contextualizing the results. This partnership makes dynamic markets more resilient, especially when there are regime shifts or other extraordinary disturbances. A paradigm shift in how businesses and analysts understand intricate, multifaceted data sets is represented by hybrid intelligence.Item GRAPH NEURAL NETWORK GNN FOR RETAIL FORECASTING: A TEMPORAL ENCODING APPROACH FOR SALES FORECASTING(2025) Safi Ullah ZahidEffective sales forecasting is a key factor in retail decision-making, allowing businesses to maximize inventory, improve supply chains, and provide superior customer experience. Traditional approaches using statistical models or machine learning fail to capture complex interdependencies among products, customers, and trends over time. In this research, we present a new forecasting framework that enriches temporal encoding using Graph Neural Networks (GNNs) to model and forecast retail sales. Leveraging Kaggle open-source sales data, the suggested model forms dynamic graphs of temporal interactions and item interactions, enabling the GNN to comprehend changing relationships over time. Experimental results showcase significant advantages when applying a GNN-based model compared to traditional methods like SARIMA and state-of-the-art deep learning architectures such as Long Short-Term Memory (LSTM) networks, achieving a markedly better prediction accuracy across various temporal horizons, especially in dynamic retail settings. Starting from the graph elaboration (the incorporation of temporal encoding) to the optimization using attention mechanisms and finally the evaluation in terms of key performance metrics (RMSE and MAE) in which evidence is given of the relevance of the approach. This method yields enhanced scalability and flexibility, especially for real-time retail forecasting use cases. The results underscore the potential of GNNs to improve forecasting accuracy, leading to better-informed business decisions and advancing the emerging area of temporal graph applications in business analyticsItem CUSTOMER CHURN PREDICTION OF PAKISTAN’S TELECOM INDUSTRY(UMT, Lahore, 2025) Nabeel AhmadIdentifying customer churn is a critical challenge in a highly competitive industry like telecom, where companies struggle to retain customers amidst market saturation, competitive pricing, and service dissatisfaction. Predicting and prevent churn is essential for the telecom providers to maintain revenue, optimize their operational costs, and enhance customer satisfaction leading to retaining of customers by turning churners to non-churners. This study aims to develop an accurate prediction model which is tailored to Pakistan’s telecom sector by leveraging machine learning techniques. This research employs an explanatory approach, a dataset of 9,760 customer records was taken from Kaggle. Key features for prediction include monthly usage, relationship duration, service type, and customer complaints. Machine learning models including Logistic Regression, Decision Trees, Random Forest, Gradient Boosting and Artificial Neural Networks (ANN)- were used and their performance was evaluated using accuracy, precision, recall, F1-score, and AUC-ROC. Results from the applied models indicate that deep learning model, ANN outperformed by achieving 97.4% accuracy, 100% recall, and 88.3% precision, which proves to be the most reliable model for churn prediction. Random forest also showed strong performance of 92.6% accuracy and 97.2% recall, balancing the interpretability with predictive power. This study highlights the importance of customer churns and also the importance of addressing the churns using SMOTE which improves model performance by making sure minority class are included. The results of research study shows that costumers who frequently use the service and have minimal relationship duration are more inclined to leave. This research offers practical insights to telecom organizations with recommendations for enhanced service quality, personalized retention methods and predictive analytics-based marketing campaignsItem Air Pollution Mitigation in Islamabad: A Data-Driven Approach Using Air Quality Index (AQI) And Climate Trends(UMT, Lahore, 2025) MUQADDAS SATTARThis study examines air pollution outcomes from an analytics framework and specifically looks at air quality index (AQI) from 2020-2024, one of the climate variables, temperature, and humidity. The primary pollutants specialist study was identified as PM2.5- and NO₂- and three models were developed using machine learning (random forest, ARIMA, and LSTM) to forecast air quality. The exploratory data analysis identified seasonal increase in pollution with climate variability, winter, and higher pollution levels due to temperature inversions. The optimal long-term prediction based AQI model was LSTM, although a kind of random forest model added a few predictor variables. This research found that combining air quality data with meteorological data did improve forecasting and potentially improved policies. The study also procured recommendations for real-time monitoring, sustainable transport, and greener public urban planning design to mitigate changes in air quality. The intent of this research was to generate a pragmatic model for operationalizing the way environmental scientists could align both predictive modelling and practice strategies into a sensible approach to better address urban air pollution challenges as demonstrated in regard to Islamabad and beyond.Item Leveraging Explainable AI for Insights and Decision-Making in Compliance Management(UMT, Lahore, 2025) Umer BaigCompliance management is an essential process across multiple industries, to ensure internal organizational standards, policies, legal and industrial regulations are followed. Traditionally the compliance management approach has gone through multiple iterations and is dependent upon manual practices which slow down the process of identifying inefficiencies, risks and bottlenecks. As industries grow more complex and require data driven decision making, organizations are opting for Artificial Intelligence (AI) and Machine Learning (ML) to ensure better compliance actions. However, one major challenge is the interpretability and lack of transparency from AI-driven decisions. This study explores the integration of Machine Learning (ML) and Explainable AI (XAI) in compliance management within a multi-sector SaaS (Software as a Service) platform. In this research we apply Logistic Regression and Random Forest to predict compliance risks based on task completion times, overdue records, priority levels, and other influencing factors. To address interpretability challenges, SHAP (Shapley Additive Explanations) and LIME (Local Interpretable Model-agnostic Explanations) are employed, ensuring stakeholders are given transparent insights from AI generated predictions. In the study we follow a structured methodology, beginning with data preprocessing, feature engineering, and exploratory data analysis (EDA) to clean compliance datasets for predictive modeling. The study finds that empirical evaluations for machine learning models can ensure effective prediction for compliance risks, with the highest predictive accuracy achieved by Random Forest.Task completion duration, overdue days, and priority levels were found to be the most influential factors in determining compliance risks. It is found that trust on AI- driven compliance predictions and acceptance is significantly improved by the deployment of Explainable AI. The study concludes that AI-powered compliance management can streamline risk assessment, optimize enforcement strategies, and enhance regulatory oversight. However, there are limitations related to data quality, model biases, and scalability must be addressed for broader adoption across industries. By demonstrating how ML and XAI can be leveraged to enhance risk prediction, regulatory transparency, and decision-making in compliance related industries, this research contributes to the growing field of AI-driven compliance. It is suggested that future work should focus on refining XAI models, better training of models, incorporating real-time AI and manual compliance monitoring, and expanding AI-driven compliance frameworks to new regulatory domains.Item Leveraging Machine Learning for Enhanced Audience Segmentation and Campaign Optimization in Facebook Advertising(UMT, Lahore, 2025) UMER PERVAIZThis research investigates the application of machine learning to enhance digital advertising campaign performance within Facebook's ecosystem, addressing current limitations in transparency and real-time optimization. Analyzing thirteen months of Facebook campaign data across diverse objectives and targeting strategies, we employed unsupervised clustering for audience categorization based on exposure patterns and supervised regression modeling for expenditure-outcome relationships. Our clustering analysis revealed four distinct audience segments, each with unique reach- frequency profiles, enabling refined retargeting, frequency management, and personalization strategies. Regression modeling quantified the relationship between advertising investment and audience reach, supporting evidence-based budget allocation. The findings demonstrate that machine learning significantly improves social media advertising efficiency by enhancing audience comprehension and predictive capabilities. This methodology provides interpretable insights for campaign optimization, addressing transparency concerns in algorithmic advertising. This study empirically validates clustering applications in digital marketing and establishes frameworks for predictive campaign planning, highlighting the potential for data-driven segmentation and forecasting while upholding ethical considerations.Item Education Content Quality Management: A Multimodal Approach(UMT, Lahore, 2025) Mishaal MaraalIn the digital age, educational content plays a crucial role in learning outcomes, yet ensuring its quality and effectiveness remains a challenge. This research presents a multimodal approach to measuring educational content quality using machine learning techniques. The study focuses on analyzing various content formats, including blogs, videos, and documents, to assess their readability, lexical complexity, and engagement levels. A dataset of 5000 URLs was collected and processed using natural language processing (NLP) techniques to extract key linguistic features. Machine learning models, including K-Means clustering, Support Vector Machines (SVM), multiple regression, and neural networks, were applied to identify patterns in content quality. The results highlight that content readability and lexical density significantly influence learner engagement. Neural networks and SVM models outperformed traditional regression methods, achieving high predictive accuracy for readability ease and lexical diversity. Findings also indicate a gap in content suitability, especially in developing regions, where students struggle with complex materials due to limited technological infrastructure. The study provides actionable insights for educational platforms, curriculum designers, and policymakers to optimize content delivery based on student learning patterns. Future research can expand the dataset, incorporate real-time OCR analysis, and integrate student performance metrics to enhance content recommendations. This research serves as a foundation for improving personalized learning experiences, ensuring that educational content is not only accessible but also effective in meeting diverse student needs