Data Science, AI and Applications: First International Conference, ICDSAIA 2025, Dhaka, Bangladesh, July 18–19, 2025, Proceedings, Part I(2681 Communications in Computer and Information Science)

About the Book

The three-volume set CCIS 2681-2683 constitutes the post-conference proceedings of the First International Conference on Data Science, Artificial Intelligence and Applications, ICDSAIA 2025, held in Dhaka, Bangladesh, during July 18-19, 2025.

The 99 full papers included in this book were carefully reviewed and selected from 190 submissions. They focus on latest advancements in data science, artificial intelligence (AI), and their applications across diverse sectors--including healthcare, education, finance, governance, agriculture, and sustainable development--highlighting its potential to solve pressing societal challenges and accelerate progress toward the Sustainable Development Goals (SDGs).

Table of Contents:
Part-vol-I. .- Deep Learning-Based Potato Leaf Disease Classification Using a Custom CNN. .- Enhanced Rice Leaf Diseases Classification Using ResNet50 on a Bangladeshi Dataset. .- AI-Driven Multilayered Cybersecurity Intelligence Framework for Critical Infrastructure Protection. .- Enhancing Candidate Selection with NLP-Driven Resume Analysis for Industry 4.0 Recruitment Systems. .- Clothes-Changing Person Re-identification with Unique Identity-Specific Attribute Details. .- MAST-GCN: Multi-Part Attention-Guided Spatial-Temporal GCN Approach for Gait-Based Person Recognition. .- Evaluation of CNNs for Flower Classification: A Study on Computational Efficiency and Model Performance. .- A Transformer based approach for Real-Time Sentiment Analysis of Transliterated Bengali Text. .- Automated Classification of Husk Species Using DenseNet121 and Vision Transformer. .- Attention-Driven Ensemble Learning: Enhancing Diabetes Prediction in Data-Scarce Environments. .- Interpretable IoT-Enabled Machine Learning Framework on Optimized Climate Information for Crop Yield and Resource Usage. .- Comprehensive Predictive Insights: Leveraging Clinical Data for Hepatitis C Prediction with Machine Learning and Deep Learning. .- A Systematic Taxonomy of Neural Network  Architectures:  Principles, Trade-offs, and Future Directions. .- Predicting and Explaining Fatal Road Casualty Types in Great Britain: A Comparative Analysis of Machine Learning, Deep Learning, and Transformers. .- Obstructive Sleep Apnea Detection using 1D CNN-LSTM approach. .- Fusing ResNet50 and VGG16 for Enhanced Diagnosis of Acute Lymphoblastic Leukemia: A MultiNet Ensemble Approach. .- Deriving Biologically Relevant Rules in Breast Cancer Subtypes using FP-Growth Algorithm. .- EEG-Based Depression Detection Using CNNs and Heatmaps. .- Deep Learning-Enhanced OCT Image Analysis Pipeline: Integrating Denoising, Super-Resolution, and Fuzzy Logic for Improved Clinical Diagnostics. .- Understanding Public Perceptions and Behaviours Towards COVID-19 Vaccination: A Multifaceted Analysis. .- Early Thyroid Disease Diagnosis Using a Hybrid Ensemble Learning Approach with Feature Selection, SMOTE, and Model Explainability. .- Preprocessed Lung Data Evaluation Using SVM for Superior Cancer Diagnosis. .- A Deep Learning Approach for Detecting Pests and Diseases in Maize Crops. .- Privacy-Preserving Prediction of Chronic Kidney Disease Using Ensemble Machine Learning with Laplacian Differential Privacy and Explainable AI. .- Channel Attention Mechanism in Hybrid Deep Learning Model for Accurate Brain Tumor Classification. .- Hybrid Artificial Intelligence for Forecasting Renewable Energy Consumption with Ensemble Machine Learning and Time Series Models. .- Optimizing American Sign Language Recognition with Binarized Neural Networks: A Comparative Study with Traditional Models. .- An Explainable and Ensemble Approach for Skin Lesion Classification Using Attention-Based Lightweight CNNs. .- Handling Imbalanced Datasets with Real-World Positive Samples in Dengue Prediction Using Machine and Deep Learning Models. .- FAKD-XAI: Feature-Aligned Knowledge Distillation with Explainable AI for Efficient Brain Tumor Classification. .- Advancing Web-based Bilingual Spam Detection System  with XLM-RoBERTa: Dataset Creation and Model Fine-Tuning. .- An Interpretable Hybrid Framework for Brain Tumor Classification: Fusion of EfficientNetV2L, ViTs, and Attention Mechanisms. .- Moment Detection at Scale: Dataset-Driven Techniques for Temporal Localization. Part-vol-II. Part-vol-III.