Abrar Faiaz

Abrar Faiaz

Auto-generated profile, based on publicly available information

Abrar Faiaz

Bangladesh University of Engineering and Technology, Dhaka, Bangladesh

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
19 December 2024
Verified
Conference Paper
Open access
AFAbrar Faiaz
Abrar Faiaz
NNNowshin Nawar
Nowshin Nawar
Early warning systems are the backbone of disaster management, but traditional systems often send static alerts with no personalized advice. This work investigates an AI-powered disaster warning system using predictive analytics and LLMs to design personalized recommendations over efficient response strategies. This system ensures that real-time information on context-specific and geographically relevant preparedness is tailored to the needs of varied populations. It was built using the LangChain framework and features an architectural composition with a predictive layer in disaster monitoring and an advisory layer for proactive safety guidelines. It integrates current weather conditions with advanced machine learning algorithms to enhance disaster prediction and utilizes LLMs in constructing customized actions, hence streamlining emergency messaging with enhanced clarity and relevance. Among the tested models, it turned out that the Gemini Pro LLM is the most effective, consistently generating outputs appropriate in context and accurate. Analysis under different temperature settings demonstrated the ability of the model to maintain high precision on various types of crises. The findings indicate that this approach can transform traditional early warning systems into more adaptive, user-centered frameworks, hence enhancing disaster management practices.

19 December 2024
Verified
Conference Paper
Open access
AFAbrar Faiaz
Abrar Faiaz
DMDipankar Mitra
Dipankar Mitra
RPRanat Das Prangon
Ranat Das Prangon
Converting any kind of data to image data can make the dataset suitable for Convolutional Neural Networks (CNNs). In this study, the NSL-KDD dataset was converted to image data using the color mapping technique and using CNN a good accuracy of 98.91% and aggregated f1 score of 0.91 was achieved. Here an image representation of each row was generated using both Hue-Saturation-Value (HSV) and Viridis colormap. Though some attack types were misclassified by the model, no attack sample of the validation dataset was classified as normal. For model training, five CNN architectures were evaluated by transfer learning from their pre-trained weights. It was found that ResNet18 performs best among all the five architectures evaluated. ResNet18 uses 1x1 convolution to reduce the number of parameters used. It means that for the classification of the colormap of this dataset complex CNN architectures are not necessary. Fine-tuning of the ResNet18, the best-performing architecture, was done using 50 epochs using an optimal learning rate. However, the accuracy mentioned above was found only after 4 epochs demonstrating good efficiency of the model training. The NSL-KDD dataset contains information about network intrusion in tabular format. Hence, this model can be used for intrusion detection purposes.

7 March 2026 in Expert Systems with Applications
Verified
Journal Article
Closed access
AFAbrar Faiaz
Abrar Faiaz
ASAlphy Shaharin
Alphy Shaharin
MRMafizur Rahman
Mafizur Rahman
Md. Bazlur RashidMd. Bazlur Rashid
Md. Bazlur Rashid
SHS.M. Quamrul Hasan
S.M. Quamrul Hasan