The 4-volume set LNCS constitutes the main proceedings of the 25th International Conference on Computational Science, ICCS 2025, which took place in Singapore, Singapore, during July 7–9, 2025. The 64 full papers and 52 short papers presented in these proceedings were carefully reviewed and selected from 162 submissions. The ICCS 2025 main track full papers are organized in volumes 15903–15905 (Parts I to III) and the ICCS 2025 main track short papers are included in volume 15906 (Part IV).

Table of Contents:
ICCS 2025 Main Track Full Papers.- Backtranslation and Paraphrasing in the LLM Era? Comparing Data Augmentation Methods for Emotion Classification.- Explainable Artificial Intelligence for Bioactivity Prediction: Unveiling the Challenges with Curated CDK2/4/6 Breast Cancer Dataset.- Bus Loop Scheduling with Dueling Double Deep Q Network.- Precise Language Deception: XAI Driven Targeted Adversarial Examples with Restricted Knowledge.- Tensorial Implementation for Robust Variational Physics-Informed Neural Networks.- Neural Parabolic Wave Equation for Refractivity Estimation.- Discover the Tractable Latent Space of Floating Offshore Wind Turbine based on a Novel GNN-Encoder-Decoder-LSTM Deep Learning Architecture.- Towards Weight-space Interpretation of Low-Rank Adapters for Diffusion Models.- Discovering Governing Equations of Geomagnetic Storm Dynamics with Symbolic Regression.- Combining Shape and Trajectory Features for Human Action Classification using a Neural Network and Synthetic Data.- Improving Object Detection Quality in Football Through Super-Resolution Techniques.- Flexible User-defined Domain Decomposition in Kilometer-Scale E3SM Land Model Simulation.- A Deeper Look into the Limitations of Early-Exit Architectures for Single and Multi-Label Classification.- Comparative Analysis of Black-Box Optimization Methods for Weather Intervention Design.- Dual Adaptive Windows Toward Concept-Drift in Online Network Intrusion Detection.- Estimating Airborne Transmission Risk for Indoor Space: Coupling Agent-based Model and Computational Fluid Dynamics.- A Dynamic Model of Customers Behavior: Integrating Econophysics and Physics-Informed Neural Networks.- Adaptive PCA-Based Outlier Detection for Multi-Feature Time Series in Space Missions.- Adaptive Physics Refinement for Anatomic Adhesive Dynamics Simulations.- Microfluidic Digital Twin for Enhanced Single-Cell Analysis.- Energy-Efficient Neural Network Training for Scientific Datasets with Advanced Similarity Analytics and Orchestration.