Fouad Trad

Mr. Fouad Trad

Electrical and Computer Engineering, American University of Beirut, Beirut, Beyrouth, Lebanon
orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
FTFouad Trad
Fouad Trad
ACAli Chehab
Ali Chehab
Large Language Models (LLMs) are reshaping the landscape of Machine Learning (ML) application development. The emergence of versatile LLMs capable of undertaking a wide array of tasks has reduced the necessity for intensive human involvement in training and maintaining ML models. Despite these advancements, a pivotal question emerges: can these generalized models negate the need for task-specific models? This study addresses this question by comparing the effectiveness of LLMs in detecting phishing URLs when utilized with prompt-engineering techniques versus when fine-tuned. Notably, we explore multiple prompt-engineering strategies for phishing URL detection and apply them to two chat models, GPT-3.5-turbo and Claude 2. In this context, the maximum result achieved was an F1-score of 92.74% by using a test set of 1000 samples. Following this, we fine-tune a range of base LLMs, including GPT-2, Bloom, Baby LLaMA, and DistilGPT-2—all primarily developed for text generation—exclusively for phishing URL detection. The fine-tuning approach culminated in a peak performance, achieving an F1-score of 97.29% and an AUC of 99.56% on the same test set, thereby outperforming existing state-of-the-art methods. These results highlight that while LLMs harnessed through prompt engineering can expedite application development processes, achieving a decent performance, they are not as effective as dedicated, task-specific LLMs.

10 May 2025 in BMC Medical Research Methodology
Verified
Journal Article
Open access
PDF file
FTFouad Trad
Fouad Trad
RYRyan Yammine
Ryan Yammine
JCJana Charafeddine
Jana Charafeddine
MCMarlene Chakhtoura
Marlene Chakhtoura
MRMaya Rahme
Maya Rahme
ACAli Chehab
Ali Chehab
Systematic reviews (SRs) are essential to formulate evidence-based guidelines but require time-consuming and costly literature screening. Large Language Models (LLMs) can be a powerful tool to expedite SRs. We conducted a comparative study to evaluate the performance of a commercial tool, Rayyan, and an in-house LLM-based system in automating the screening of a completed SR on Vitamin D and falls. The SR retrieved 14,439 articles, and Rayyan was trained with 2,000 manually screened articles to categorize the rest as most likely to exclude/include, likely to exclude/include and undecided. We analyzed Rayyan’s title/abstract screening performance using different inclusion thresholds. For the LLM, we used prompt engineering for title/abstract screening and Retrieval-Augmented Generation (RAG) for full-text screening. We evaluated performance using article exclusion rate (AER), false negative rate (FNR), specificity, positive predictive value (PPV), and negative predictive value (NPV). Additionally, we compared the time required to complete screening steps of the SR using both approaches against the manual screening method. Using Rayyan, including considered as undecided or likely to include for title/abstract screening resulted in an AER of 72.1% and an FNR of 5%. The total estimated screening time, including manual review of articles flagged by Rayyan, was 54.7 hours. Lowering the Rayyan threshold to ‘likely to exclude’ reduced the FNR to 0% and the AER to 50.7%, but increased the screening time to 81.3 h. Using the LLM system, after title/abstract and full-text screening, 78 articles remained for manual review, including all 20 identified by traditional methods. The LLM achieved an AER of 99.5%, specificity of 99.6%, PPV of 25.6%, and NPV of 100%, with a total screening time of 25.5 h, including manual review of the 78 articles, reducing the manual screening time by 95.5%. The LLM-based system significantly enhances SR efficiency, compared to manual methods and Rayyan while maintaining low FNR.

16 November 2024 in Computers in Biology and Medicine
Verified
Journal Article
Open access
ZKZeina El Kojok
Zeina El Kojok
HKHadi Al Khansa
Hadi Al Khansa
FTFouad Trad
Fouad Trad
ACAli Chehab
Ali Chehab
In recent years, deep learning has become a popular tool to analyze and classify medical images. However, challenges such as limited data availability, high labeling costs, and privacy concerns remain significant obstacles. As such, generative models have been extensively explored as a solution to generate new images and overcome the stated challenges. In this paper, we augment a dataset of chest CT scans for Vertebral Compression Fractures (VCFs) collected from the American University of Beirut Medical Center (AUBMC), specifically targeting the detection of incidental fractures that are often overlooked in routine chest CTs, as these scans are not typically focused on spinal analysis. Our goal is to enhance AI systems to enable automated early detection of such incidental fractures, addressing a critical healthcare gap and leading to improved patient outcomes by catching fractures that might otherwise go undiagnosed. We first generate a synthetic dataset based on the segmented CTSpine1K dataset to simulate real grayscale data that aligns with our specific scenario. Then, we use this generated data to evaluate the generative capabilities of Deep Convolutional Generative Adverserial Networks (DCGANs), variational autoencoders (VAEs), and VAE-GAN models. The VAE-GAN model demonstrated the highest performance, achieving a Fréchet Inception Distance (FID) five times lower than the other architectures. To adapt this model to real-image scenarios, we perform transfer learning on the GAN, training it with the real dataset collected from AUBMC and generating additional samples. Finally, we train a CNN using augmented datasets that include both real and generated synthetic data and compare its performance to training on real data alone. We then evaluate the model exclusively on a test set composed of real images to assess the effect of the generated data on real-world performance. We find that training on augmented datasets significantly improves the classification accuracy on a test set composed of real images by 16 %, increasing it from 73 % to 89 %. This improvement demonstrates that the generated data is of high quality and enhances the model's ability to perform well against unseen, real data.


23 July 2025 in ACM Transactions on Intelligent Systems and Technology
Verified
Journal Article
Open access
FTFouad Trad
Fouad Trad
ACAli Chehab
Ali Chehab
The success of Large Language Models (LLMs) has spurred the rise of Large Multimodal Models (LMMs), which integrate multiple modalities, such as text and images, to address complex data analysis tasks. As these black-box models gain popularity due to their ease of use and adaptability, there is growing interest in understanding their potential to replace or complement task-specific models in domain-specific applications. This article evaluates the applicability and effectiveness of prompt-engineered LMMs, specifically LLaVA, BakLLaVA, Moondream, Gemini 1.5 Flash, and GPT-4o, compared to fine-tuned Vision Transformer (ViT) models in addressing cybersecurity challenges of varying complexity. Our study examines three distinct tasks: (1) detecting visual triggers indicative of potential backdoors in two scenarios (digit recognition and traffic sign classification), (2) identifying phishing attempts from Web site screenshots, and (3) classifying malware based on visual representations. The results reveal that prompt-engineered LMMs perform competitively on tasks with visually evident or moderately complex features, such as trigger detection and phishing classification, with GPT-4o and Gemini 1.5 Flash demonstrating superior performance among LMMs. However, for the highly specialized task of malware classification, LMMs exhibit notable limitations when relying solely on prompting. Fine-tuning GPT-4o significantly improves its performance, yet it still lags behind fine-tuned ViT models, which consistently achieve higher accuracy across all tasks. While ViTs deliver superior precision and robustness, they require substantial resources for training, fine-tuning, and maintenance. In contrast, LMMs provide flexibility and ease of deployment, making them an appealing alternative for scenarios where resource constraints or rapid implementation are critical. This study highlights the tradeoffs between these approaches, emphasizing that while ViTs are indispensable for high precision, specialized applications, LMMs offer a scalable and versatile solution for less complex or resource-limited tasks.