Advances in Computational Intelligence

The two-volume set LNCS 16008 & 16009 constitutes the refereed conference
proceedings of the 18th International Work-Conference on Advances in Computational Intelligence, IWANN 2025, held in A Coruña, Spain, during June 1618, 2025.

The 103 revised full papers presented in these proceedings were carefully reviewed and selected from 144 submissions. The papers are organized in the following topical sections:

Part I: Advanced Topics in Computational Intelligence; AI:Bioinformatics and Biomedical Applications; ANN HW-Accelerators; Bio-Inspired Systems and Neuro-Engineering; Recent Advances in Deep Learning; Deep Learning Applied to Computer Vision, Healthcare and Robotics; and Emerging Methodologies in Time Series Forecasting.

Part II: Explainable and Interpretable Machine Learning (xAI) with a Focus on Applications; General Applications of AI; ITOMAD Intelligent Techniques for Optimization, Modeling, and Anomaly Detection; Machine Learning for 4.0 Industry Solutions; Machine Learning for Photovoltaic System Optimization and Control in Modern Energy Grids; New and future advances in BCI-based Spellers; and Social and Ethical aspects of AI.

Klappentext

.- Advanced Topics in Computational Intelligence. .- Power Quality 24-hour Prediction Based on L-Transform Derivative Modular and Deep Learning Statistics Using Environmental Data in detached Smart Buildings. .- Incremental Feature Learning of Shallow Feedforward Regression Neural Networks using Particle Swarm Optimisation. .- Resilience Under Attack: Benchmarking Optimizers Against Poisoning in Federated Learning for Image Classification Using CNN. .- VIDEM: VIDeo Effectiveness and Memorability Dataset. .- Penetration Testing with AI: Case Studies on LLM and RL-Based Attack Agents. .- A comparative study of deep learning approaches for classifying wild and cultivated fish. .- Sparse Least Square SVM in Primal via Nesterov Accelerated Alternating Directions Method of Multipliers. .- AI:Bioinformatics and Biomedical Applications. .- A transformer-based model to predict micro RNA interactions. .- Leveraging Large Language Models on Assay Descriptions to Improve the Prediction of Inhibitors for Mycobacterium tuberculosis. .- Advancing Imminent Fracture Risk Prediction: Integrating Machine Learning with Enhanced Feature Engineering. .- Self-organizing Maps for Missing Value Imputation in Transcriptomic Datasets. .- ANN HW-Accelerators. .- RECS: A Scalable Platform for Heterogeneous AI Acceleration in the Cloud-Edge Continuum. .- Evaluating HBM to accelerate neural networks on FPGAs demonstrated using binary neural associative memories. .- Resource-efficient Implementation of Convolutional Neural Networks on FPGAs with STANN. .- High-Performance FPGA-based CNN Acceleration for Real-Time DC Arc Fault Detection. .- Optimizing AI on the Edge: Partitioning Neural Networks Across Heterogeneous Accelerators. .- Comparison of Hardware Component and Manycore Implementation for Anomaly Detection in Trustworthy System-on-Chips. .- Bio-Inspired Systems and Neuro-Engineering. .- An Emotional Classifier for Machine’s Artificial Visual Aesthetic Appraisal. .- Hardware and Software influence on EAs power consumption. .- Properties of monoclinic gallium oxide film and its photomemristor application in nonlinear RMC circuit. .- A perceptron-like neural network implementing a learning-capable K-nearest neighbor classifier. .- From Biological Neurons to Artificial Neural Networks: A Bioinspired Training Alternative. .- Recent Advances in Deep Learning. .- Domain Adaptation of the Whisper ASR Model for Tourism Call Center Transcription in Polish. .- Learning to Search with Subgoals. .- Towards Speaker Independent Speech Emotion Recognition by means of Dataset Aggregation. .- Learning Heuristics for k-NANN-A*: A Deep Learning Approach. .- Evaluating Higher-Level and Symbolic Features in Deep Learning on Time Series: Towards Simpler Explainability. .- Energy-Efficient Radio Resource Allocation in 5G Using Deep Q-Networks. .- Multi-view Cross Contrastive Learning for Multimodal Knowledge Graph Recommendation. .- MuleTrack: A Lightweight Temporal Learning Framework for Money Mule Detection in Digital Payments. .- Modular Deep Neural Networks with residual connections for predicting the pathogenicity of genetic variants in non coding genomic regions. .- Modeling Student–Subject Interactions with GNNs for Grade Prediction. .- Deploying Vision Foundation AI Models on the Edge. The SAM2 Experience. .- Generative AI for Contextualizing Bronze Age Objects in Historical Scenes. .- G-TED SAM: Node Classification via Graph Transformer to Simple Attention Model Distillation. .- Expression Recognition in Faces Partially Occluded by Head-Mounted Displays. .- Reinforcement Learning for Mapless Navigation: Enhancing Exploration with Image-Based Rewards

Zusammenfassung

The two-volume set LNCS 16008 & 16009 constitutes the refereed conference
proceedings of the 18th International Work-Conference on Advances in Computational Intelligence, IWANN 2025, held in A Coruña, Spain, during June 1618, 2025.

The 103 revised full papers presented in these proceedings were carefully reviewed and selected from 144 submissions. The papers are organized in the following topical sections:

Part I: Advanced Topics in Computational Intelligence; AI:Bioinformatics and Biomedical Applications; ANN HW-Accelerators; Bio-Inspired Systems and Neuro-Engineering; Recent Advances in Deep Learning; Deep Learning Applied to Computer Vision, Healthcare and Robotics; and Emerging Methodologies in Time Series Forecasting.

Part II: Explainable and Interpretable Machine Learning (xAI) with a Focus on Applications; General Applications of AI; ITOMAD Intelligent Techniques for Optimization, Modeling, and Anomaly Detection; Machine Learning for 4.0 Industry Solutions; Machine Learning for Photovoltaic System Optimization and Control in Modern Energy Grids; New and future advances in BCI-based Spellers; and Social and Ethical aspects of AI.

Inhalt

.- Advanced Topics in Computational Intelligence .

.- Power Quality 24-hour Prediction Based on L-Transform Derivative Modular and Deep Learning Statistics Using Environmental Data in detached Smart Buildings.

.- Incremental Feature Learning of Shallow Feedforward Regression Neural Networks using Particle Swarm Optimisation.

.- Resilience Under Attack: Benchmarking Optimizers Against Poisoning in Federated Learning for Image Classification Using CNN.

.- VIDEM: VIDeo Effectiveness and Memorability Dataset.

.- Penetration Testing with AI: Case Studies on LLM and RL-Based Attack Agents.

.- A comparative study of deep learning approaches for classifying wild and cultivated fish.

.- Sparse Least Square SVM in Primal via Nesterov Accelerated Alternating Directions Method of Multipliers.

.- AI:Bioinformatics and Biomedical Applications .

.- A transformer-based model to predict micro RNA interactions.

.- Leveraging Large Language Models on Assay Descriptions to Improve the Prediction of Inhibitors for Mycobacterium tuberculosis.

.- Advancing Imminent Fracture Risk Prediction: Integrating Machine Learning with Enhanced Feature Engineering.

.- Self-organizing Maps for Missing Value Imputation in Transcriptomic Datasets.

.- ANN HW-Accelerators .

.- RECS: A Scalable Platform for Heterogeneous AI Acceleration in the Cloud-Edge Continuum.

.- Evaluating HBM to accelerate neural networks on FPGAs demonstrated using binary neural associative memories.

.- Resource-efficient Implementation of Convolutional Neural Networks on FPGAs with STANN.

.- High-Performance FPGA-based CNN Acceleration for Real-Time DC Arc Fault Detection.

.- Optimizing AI on the Edge: Partitioning Neural Networks Across Heterogeneous Accelerators.

.- Comparison of Hardware Component and Manycore Implementation for Anomaly Detection in Trustworthy Sys…