http://hdl.handle.net/123456789/10358 Wed, 05 Aug 2026 13:39:09 GMT 2026-08-05T13:39:09Z http://hdl.handle.net/123456789/19220 Mobile Network Traffic Classification and Prediction Hassan Ayaz, 01-241222-010 Call Detail Records (CDRs) from mobile networks offer rich insights into network performance and user behavior. In this study, we analyze CDR data from Telecom Italia, encompassing spatiotemporal patterns across Milan, segmented into a 100x100 grid with each cell covering 0.3 kilometers. By analyzing the spatiotemporal dynamics of CDR data, we classify the network traffic into four categories: highest, high, moderate and low with high network traffic regions predominantly located in the city center. After network traffic classification, we predict future traffic patterns. We utilize automated machine learning (AutoML) tools and the state-of-the-art TimeGPT model for network traffic forecasting. Comparative analysis reveals that AUTOML performs better then TIMEGPT, delivering superior prediction and performing better on the various evaluation metrices resultantly capturing complex temporal and spatial relationships in the data. These predictive capabilities enable dynamic resource allocation, enhanced congestion management and improved network efficiency. Our findings underscore the potential of both AUTOML and TimeGPT, to some extent AUTOML appears to be more scalable and adaptable solution for mobile network traffic classification and forecasting, marking a significant advancement in network planning and optimization for urban environments Supervised by Dr. Kashif Sultan Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/19220 2025-01-01T00:00:00Z http://hdl.handle.net/123456789/20458 Psychological Disord Ers Diagnosis Framework Using Reflective Listening and Generative AI Mahrukh Shakoor, 09-241241-001 Mental health disorders such as Generalized Anxiety Disorder (GAD), Major Depressive Disorder (MDD), and Borderline Personality Disorder (BPD) are rising globally, and this causes access to clinically trained professionals limited. To address this gap, our research proposes an innovative Psychological Disorder Diagnosis Framework Using Reflective Listening and Generative AI, integrating multimodal, natural language processing, DSM-5 diagnostic logic, knowledge-graph based reasoning, and validated psychometric assessments under one system. The system conducts a multi-stage evaluation process that starts with two rounds of conversational screening, followed by NLP-driven symptom extraction, keyword clustering, probability scoring, and DSM-5 validation checks. The reflective listening technique is used in fine-tuned LLM-based empathetic dialogue to enhance emotional understanding and user comfort. A secondary diagnostic stage executes standardized clinical tests, including the Penn State Worry Questionnaire (PSWQ) for GAD, Beck Depression Inventory (BDI) for MDD, and McLean Screening Instrument (MSI) for BPD, ensuring that AI predictions are verified against clinical standards. A comprehensive Gold dataset was created through system-generated session logs, enhanced with DSM-5 ground-truth labels, validated psychometric scores, corrected knowledge-graph structures, and reflective-listening quality ratings provided by licensed psychologists. This dataset supports the fine-tuning of a generative model capable of producing empathetic reflections, accurate disorder predictions, DSM-5 aligned reasoning, and appropriate treatment suggestions. The architecture combines a user-friendly multilingual interface, text and audio inputs, an NLP and reflective-listening engine, a clinical rule-based inference module, and a visualization layer that provides probability scores, knowledge graphs, and diagnostic summaries. Results determine that the proposed framework enhances diagnostic transparency, cultural adaptability, and clinical validity compared to traditional sentiment-based mental health models. The system achieves high stability between AI predictions and clinician evaluations, while reflective listening significantly improves user engagement and emotional coherence. This research contributes an explainable, empathetic, and clinically grounded AI diagnostic framework, establishing a foundation for next-generation intelligent mental health assessment systems suitable for real-world psychological support and early screening applications. Supervised by  Dr. Tamim Ahmad Khan Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/20458 2025-01-01T00:00:00Z http://hdl.handle.net/123456789/21538 Self-Healing And Lightweight Intrusion Detection For Iot-Based Systems Mahawish, 02-200182-006 The rapid growth of Internet of Things (IoT) has revolutionized modern technology, enabling seamless connectivity across diverse domains. However, resourceconstrained IoT devices are highly vulnerable to cyber threats, particularly Denial of Service (DoS) and Distributed Denial of Service (DDoS) attacks. Such attacks can severely compromise the reliability and functionality of IoT networks, necessitating robust, resource-efcient, and adaptive intrusion detection mechanisms. Traditional Intrusion Detection Systems (IDSs) often fail to meet the unique demands of IoT environments due to their high computational overhead and inability to adapt dynamically to emerging threats. Addressing these challenges requires an innovative approach that balances lightweight design with high detection accuracy and resilience. The frst part of this research proposes an Ensemble Lightweight Intrusion Detection System (ELIDS) approach, which combines the capabilities of seven distinct flter-based Feature Selection (FS) methods to identify the most relevant features in classifying normal and DoS/DDoS attack packets. In the proposed ELIDS, each FS method generates a ranked list of features along with their corresponding scores. These lists are subsequently aggregated, resulting in a consolidated fnal list that includes the reduced set of features. The selected features are further utilized to train six Machine Learning (ML) algorithms, designing lightweight ML-enabled IDS. The proposed ELIDS is evaluated for in-domain and cross-domain testing scenarios. The results reveal that incorporating an ensemble FS approach improves detection accuracy also optimizes computational resources. Moreover, ELIDS outperforms conventional FS methods, especially over cross-domain testing scenarios. The second part of this research presents a Self-Healing for Internet of Thing (SHviii IoT), designed to enhance the performance of ML-enabled IDS models, especially when a decline in their effectiveness is observed during cross-domain testing. In this framework, the health monitor continuously checks the behavior of simulated IoT devices and generates a danger signal when it detects any deviation from normal operating patterns caused by a DoS/DDoS attack. Once the danger signal is triggered, the proposed SHIoT initiates its healing mechanism, which activates defensive measures to counter DoS/DDoS attacks. The framework is assessed using performance metrics and obtained results demonstrate a notable enhancement in the performance of ML models, which retain both previous and newly acquired knowledge to identify both existing and emerging DoS/DDoS attacks. Finally, the third part of this research integrates the Self-Healing and Ensemble-based Lightweight IDS (SHIoT-ELIDS). The SHIoT-ELIDS is evaluated using nine unseen datasets, achieving a peak accuracy of 99.9% while employing Random Forest as the learning algorithm. Additionally, the proposed SHIoT-ELIDS performs efciently in terms of resource utilization, where testing a single packet takes only 0.0050 msec, demonstrating its appropriateness for real-time detection environments. The CPU usage is around 0.0014% per packet, ensuring the system imposes minimal load on the underlying hardware. Additionally, memory consumption remains low at 0.1456 MB per packet, while the fnal classifcation model size is of 697 KB, making it ideal for deployment on devices with limited processing and storage capacity, such as IoT devices. Overall, SHIoT-ELIDS can improve the security posture of IoT networks, providing a scalable and resource-efcient defense against evolving cyber threats Supervised by Dr. Osama Rehman Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/21538 2025-01-01T00:00:00Z http://hdl.handle.net/123456789/21539 An Agile Approach for Inquiry-Based Learning in Ubiquitous Environment Bushra Fazal Khan, 02-200181-009 Education is an evolving field where challenges are constantly changing of content and learners' needs. Thus, there was a shift in pedagogy from teacher-centric to student-centric learning with the advancement of technology. In recent times, the concept of smart classrooms or ubiquitous learning (u-learning) environment has been prevalent in the learning domain as it increases the accessibility of resources and learning activities anywhere, anytime using smart devices and internet. Also, enhance engagement by creating interactive, engaging, and enjoyable learning experience. Inquiry-based learning (IBL) is an effective instructional strategy that can be in the form of a problem or task used for triggering student engagement and enhancing students’ knowledge by introducing inquiring activities for fostering critical thinking, problem-solving, and student engagement. Also, ubiquitous learning environments integrate real-world and digital-world resources and learning scenarios, thus providing a new opportunity for implementing technologyenhanced inquiry-based activities to enrich student learning experience. But, implementing IBL in a ubiquitous learning environment within school settings presents its own set of challenges, especially in structuring and controlling collaboration effectively. Collaborative learning is a process of sharing knowledge and constructing meaning by complete involvement of all the peers discuss what they know and gain more knowledge with the help of other peers. Agile methodologies are very successful in the software engineering domain for collaboration professional development environment. These methodologies have been used for quite some time to teach software engineering concepts and agile development using a projectdriven approach in computer science. Due to their incremental, iterative, adaptive, and collaborative nature, they have gained popularity in educational domains. There is a lack in literature about the use of agile methods in IBL in ubiquitous environment. Moreover, there is also the opportunity to develop agile-based teaching approaches for other disciplines other than computing, particularly for school students.VIII The significance of this study is about the developing and evaluating the ScrumBan Ubiquitous Inquiry Framework (SBUIF) through the uASK application for collaborative IBL amongst school students. For evaluation purposes, computer-supported collaborative learning (CSCL) affordances, along with micro and meso levels of the M3 evaluation framework, have been applied. An experiment was conducted to test the effectiveness of the uASK application against Trello application. The sample involved 205 seventh-grade students. Results showed that the uASK learners scored better than those using Trello. At the micro-level, student satisfaction score by uASK users was 77.14% in comparison to Trello users which was 54.12%. It showed significant improvement and revealing better results in technology, communication, and learning performance. The meso-level measures of students learning performance revealed that uASK users had better learning outcomes then Trello (68.88% vs 58.13% for post-test improvement). Also more efficient use of time (34.7% faster). Thus, the results showed that uASK users were more engaged, satisfied, and enjoyed uASK more than the other users. The study further recommends that uASK is more effective than Trello in promoting collaborative learning activities inquiring approach in any place. The study demonstrates that the use of such framework and application based on SBUIF can support agile based collaborative inquiry learning at school level, thereby advancing knowledge in the fields of agile and inquiry-based education. Furthermore, by implementing the framework in a real-world classroom setting, it highlights its practical value, offering guidance for educators in designing similar applications to support future teaching and learning Supervised by Dr. Sohaib Ahmed Wed, 01 Jan 2025 00:00:00 GMT http://hdl.handle.net/123456789/21539 2025-01-01T00:00:00Z