Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Muhammad Faraz Manzoor has a total of 26 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Redefining Independent Learning With AI Agents

    by Junaid Nasir Qureshi, Muhammad Faraz Manzoor

    Self-guided AI agents are rapidly shifting education from static instruction to adaptive, learner-controlled systems operating at scale. These agents personalize learning, provide continuous support, and assist with complex tasks, but they also complicate identity verification, skill attestation, and attribution of human versus AI-generated work. This threatens trust in independently acquired, AI-supported credentials. This chapter presents a standards-based trust framework that combines World Wide Web Consortium (W3C) Decentralized Identifiers (DID) and Verifiable Credentials (VC) with content provenance specifications from the Coalition for Content Provenance and Authenticity (C2PA). We propose a four-layer architecture Identity, Trust, Data, and Application supported by a reference blueprint covering credential issuance, revocation, selective disclosure, and provenance verification. The design separates off-ledger personal data storage from on-ledger cryptographic proofs to preserve privacy while enabling credential portability, verifiable assessments.

  • Open Access

    Secure and scalable federated learning for citrus disease detection with CNN-ANN and data privacy for agriculture

    by Abdullah Mehboob, Muhammad Shoaib Farooq, Muhammad Faraz Manzoor, Momina Shaheen, Junaid Nasir Qureshi

    1. Citrus diseases, including citrus canker, greening, and black spots, present a severe threat to global agriculture, particularly to citrus farming, which is important for thousands of livelihood...

  • Exploring the Shift From Traditional Education Models to AI-Supported Human Collaboration

    by Muhammad Faraz Manzoor, Daksh Dave

    This chapter reviews how artificial intelligence is reshaping education by moving traditional teacher-centered systems toward more personalized, creative, and inquiry-driven learning. It outlines the historical shift from conventional instruction to AI-supported models that use adaptive pathways, competency-based structures, collaborative learning, and continuous assessment. The discussion explains how teacher responsibilities are expanding toward roles as learning designers, AI supervisors, and ethical stewards, while students gain greater autonomy, agency, and metacognitive skills. Key ethical and policy concerns, including privacy, bias, accessibility, and governance, are addressed to guide responsible adoption. Global cases show varied approaches and offer lessons for effective integration. The chapter closes with a view of emerging AI-native ecosystems featuring hybrid learning spaces, lifelong AI assistance, and analytics-driven credentialing aimed at building a more equitable and human-centered educational future.

  • Open Access

    OSSAPTestingPlus: A Blockchain-Based Collaborative Framework for Enhancing Trust and Integrity in Distributed Agile Testing of Archaeological Photogrammetry Open-Source Software

    by Omer Aziz, Muhammad Shoaib Farooq, Junaid Nasir Qureshi, Muhammad Faraz Manzoor, Momina Shaheen

    (1) Background: A blockchain-based framework for distributed agile Open-Source Software for Archaeological Photogrammetry (OSSAP) testing life cycle is an innovative approach that uses blockchain technology to optimize the Open-Source Software for Archaeological Photogrammetry process. Previously, various methods have been employed to address communication and collaboration challenges in Open-Source Software for Archaeological Photogrammetry, but they were inadequate in aspects such as trust, traceability, and security. Additionally, a significant cause of project failure was the non-completion of unit testing by developers, leading to delayed testing. (2) Methods: This article discusses the integration of blockchain technology in Open-Source Software for Archaeological Photogrammetry and resolves critical concerns related to transparency, trust, coordination, testing and communication. A novel approach is proposed based on a blockchain framework named Open-Source Software for Archaeological Photogrammetry Testing-Plus. (3) Results: The Open-Source Software for Archaeological Photogrammetry Testing-Plus framework utilizes blockchain technology to provide a secure and transparent platform for acceptance testing and payment verification. Moreover, by leveraging smart contracts on a private Ethereum blockchain, Open-Source Software for Archaeological Photogrammetry Testing-Plus ensures that both the testing team and the development team are working towards a common goal and are compensated fairly for their contributions. (4) Conclusions: The experimental results conclusively show that this innovative approach substantially improves transparency, trust, coordination, testing and communication and provides security for both the testing team and the development team engaged in the distributed agile Open-Source Software for Archaeological Photogrammetry (Open-Source Software for Archaeological Photogrammetry) testing life cycle.


  • Open Access

    Energy Load Forecasting with Machine Learning: Models, Metrics, and Future Directions

    by Muhammad Faraz Manzoor

    Energy load forecasting plays a crucial role in the efficient management and operation of smart grids, enabling utilities to optimize energy distribution and improve grid reliability. Recent advancements in machine learning (ML) techniques have significantly enhanced the accuracy and adaptability of energy load prediction models. This review explores various ML models used for energy load forecasting, including traditional models, deep learning approaches such as long short-term memory (LSTM) networks and gated recurrent units (GRUs), and ensemble methods. The review discusses the strengths and limitations of each approach, highlighting its applicability to different forecasting timeframes and data characteristics. Additionally, it examines the performance evaluation metrics commonly used to assess model accuracy and reliability. Although multiple studies have examined forecasting techniques, there is a lack of comprehensive evaluation that connects model choice with practical deployment constraints such as data quality, real-time scalability, and interpretability. This review addresses this gap by systematically analyzing the challenges and emerging solutions in the context of smart grid applications. Despite the progress made, challenges related to data quality, computational complexity, and model interpretability remain significant barriers. The review concludes with an exploration of emerging trends and future directions in energy load forecasting, including hybrid models, federated learning, and reinforcement learning, which offer promising solutions to overcome existing limitations and improve forecasting performance in smart grid systems. Overall, the findings suggest that while no single model is universally optimal, integrating external factors, improving data quality, and adopting hybrid or explainable artificial intelligence (AI) approaches are critical for building more accurate, scalable, and interpretable forecasting systems.


  • Open Access

    Detecting Fake News in Urdu Language Using Machine Learning, Deep Learning, and Large Language Model-Based Approaches

    by Muhammad Shoaib Farooq, Syed Muhammad Asadullah Gilani, Muhammad Faraz Manzoor, Momina Shaheen

    Fake news is false or misleading information that looks like real news and spreads through traditional and social media. It has a big impact on our social lives, especially in politics. In Pakistan, where Urdu is the main language, finding fake news in Urdu is difficult because there are not many effective systems for this. This study aims to solve this problem by creating a detailed process and training models using machine learning, deep learning, and large language models (LLMs). The research uses methods that look at the features of documents and classes to detect fake news in Urdu. Different models were tested, including machine learning models like Naïve Bayes and Support Vector Machine (SVM), as well as deep learning models like Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM), which used embedding techniques. The study also used advanced models like BERT and GPT to improve the detection process. These models were first evaluated on the Bend-the-Truth dataset, where CNN achieved an F1 score of 72%, Naïve Bayes scored 78%, and the BERT Transformer achieved the highest F1 score of 79% on Bend the Truth dataset. To further validate the approach, the models were tested on a more diverse dataset, Ax-to-Grind, where both SVM and LSTM achieved an F1 score of 89%, while BERT outperformed them with an F1 score of 93%.


  • Open Access

    AI‐Driven Adaptive Federated Learning With Privacy Preservation and Imbalance Adjustment for Financial Credit Card Fraud Detection

    by Muhammad Shoaib Farooq, Shahzada Fahad Munir, Muhammad Faraz Manzoor, Momina Shaheen

    The increasing complexity of fraudulent activities requires advanced fraud detection systems, as existing solutions lack effectiveness due to two challenges. First, privacy concerns prevent financial institutions from sharing sensitive transaction data. Second, data imbalance causes biased models, as fraudulent transactions represent a small fraction of total transactions, leading to poor fraud detection performance. To address these challenges, we propose an AI‐driven adaptive federated learning (AFL) framework for credit card fraud detection (CCFD). AFL enables decentralized learning, allowing financial institutions to train a global fraud detection model collaboratively without sharing raw transaction data. The model aggregation is performance‐adaptive, weighting client contributions based on detection accuracy to ensure a robust global model. To overcome data imbalance, we introduce a multistep data balancing framework integrating Tomek links for undersampling, borderline‐SMOTE for oversampling, and cognitive sample pruning to remove misleading samples. To evaluate the robustness and generalizability of the proposed framework, we conducted experiments on both the widely used 2013 Kaggle dataset and the Sparkov simulated dataset (2019‐2020). The Sparkov dataset, which contains interpretable demographic and merchant‐level features, allowed us to test the model’s adaptability to diverse data sources. The results demonstrate that the proposed AFL framework, combined with advanced data balancing, significantly outperforms traditional models, achieving 99% accuracy, 99.5% precision, 99.4% recall, and 99% F1‐score on the Kaggle dataset, and 97.4% accuracy, 99.5% precision, 97.5% recall, and 98.4% F1‐score on the Sparkov dataset. This research highlights AI’s transformative role in finance, particularly in enhancing fraud detection systems with improved accuracy, robustness, security, and scalability.


  • Open Access

    A Review of Machine Learning Techniques for Precision Agriculture and Crop Yield Prediction

    by Muhammad Faraz Manzoor

    Agriculture plays a critical role in the economic development of nations. However, with a growing global population, shifting climate conditions, and limited natural resources, meeting the increasing food demand is becoming a complex challenge. Precision agriculture, also referred to as smart farming, has emerged as a powerful solution to address these issues by promoting sustainable agricultural practices. At the heart of this innovation is machine learning (ML), ensemble learning, and deep learning, which enable systems to learn and adapt without explicit programming. Together with IoT-enabled farm equipment, ML is driving the next agricultural revolution. In this systematic literature review (SLR), we examine research from 2015 to 2023 on the diverse applications of ML in agriculture. A key area explored in this SLR is disease detection, focusing on identifying diseases. Yield prediction is another prominent area, where ML is used to forecast crop yields based on diverse environmental and agricultural data. The review also examines the prediction of soil properties, such as organic carbon content and moisture levels, which are crucial for optimizing crop management. Lastly, weather prediction is discussed, highlighting its importance in planning and adjusting farming activities in response to climatic conditions. Additionally, challenges such as data availability, model scalability, and the need for regional validation are explored.