Code, trained models and metadata accompanying the paper Optimizing Spectrogram Front-Ends and Training Strategies for Deep Learning-Based Acoustic Insect Classification (Peter & Kunzelmann, submitted). The models classify insect sounds (451 species of Orthoptera and Cicadidae) and are trained on InsectSet459 (Faiß & Stowell, 2025). Contents Code ( src/ , scripts/ , configs/ ): data download and preparation, training and evaluation (PyTorch, EfficientNetV2-S). Trained models ( models/ ): best_model.pth (64 mel bands, mixup, class-weighted loss) and baseline_model.pth (128 mel bands, no mixup, no class weighting). The checkpoints contain the m
Peter, Tobias J., Kunzelmann, Jaromir
Overview This repository provides the source code, experimental configurations, and supporting reproducibility materials for the research study titled "An Explainable Stacked Ensemble Framework for Multi-Horizon Temperature Prediction in IoT-Enabled Medical Cold Chain Systems." The study investigates OPTIML, a multi-horizon stacked ensemble framework for indoor temperature forecasting using environmental measurements collected from an IoT-enabled medical cold-chain monitoring system. The proposed OPTIML framework integrates ten machine learning and deep learning base learners with a Lasso-based meta-learner to generate temperature forecasts a
Ranga, Surbhi, Dutt, Varun, Rao, Akash, Venkata Uday, Kala
This record publishes the DIO/ODI Reference Model v0.6.0 , a versioned machine-readable reference layer for Digital Identity Optimization (DIO) and the Ontology of Digital Identity (ODI) . DIO is the applied discipline concerned with improving the conditions under which an entity's digital identity is reconstructed from its distributed digital representational state; ODI provides the theoretical meta-layer for examining the existence, persistence, variation, and reconstruction of that identity. The model is designed to make the DIO/ODI conceptual system consistently identifiable and interpretable across heterogeneous environments including ar
Beránek, Daniel
OmicsVisStat (v1.0.1) This maintenance release includes bug fixes, stability improvements, and minor refinements to enhance the overall performance and usability of OmicsVisStat . No major changes have been made to the core statistical analysis, machine learning, visualization, or data processing workflows introduced in Version 1.0. Highlights Fixed minor bugs reported in Version 1.0 Improved software stability and performance Refined user interface and workflow behavior Updated documentation and package metadata Minor code optimizations and maintenance improvements Availability Updated standalone executable and source code Zenodo archive upd
Daramola, Oluwatosin, Mechref, Yehia
Code and data accompanying the paper "Prediction accuracy and optimization outcome diverge with dimensionality: surrogate selection for metaheuristic design of code-checked frames" (submitted to Engineering Applications of Artificial Intelligence). The archive contains two self-contained frame design benchmarks (a 30-variable reinforced concrete frame checked to ACI 318 and a 60-variable steel frame checked to AISC 360), four surrogate models (Kriging, a feedforward neural network, XGBoost and radial basis interpolation) behind one interface, a real-coded genetic algorithm with Deb ranking and a repair operator, the results of all 800 optimiz
Seephandorn, Yodchai, Musiket, Kamtornkiat, Pongsivasathit, Supasit, Phungpaingam, Boonchai, Li, Kai, Lipiloet, Sukhom
Reproducibility package for “Intelligent fault detection in solar photovoltaic panels through thermal imaging and machine learning”. Contents: • Analysis pipeline: image-grouped nested cross-validation (5 outer × 3 inner folds, 20 repetitions) with Optuna hyperparameter search, HOG features and SVM, KNN, random forest and decision tree classifiers • Patch-extraction script with healthy patches sampled to match the hotspot size distribution, and the patch metadata (coordinates, class, size, source image) • Comparison descriptors: dense SIFT and SURF bags of visual words, KAZE at detected keypoints, LBP, GLCM, Lab* colour statistics, raw pixels
Chande, Haresh, Jethva, Harikrishna, Chan, Choon Kit, Gochoo, Munkhjargal, Sonawane, Chandrakant, Ketan, Kotecha
Code and data of the article "Cobetia optimization algorithm: a new bio-inspired metaheuristic for continuous problems", submitted to Artificial Intelligence Review. The record holds three files. COA_code.zip contains COA in Python and MATLAB, the implementations of the 31 competitors, the experiment runner, and the statistical analysis. COA_MATLAB.zip contains the MATLAB version of COA alone, with an example on the sphere function. COA_data.zip contains the raw result of every run of every study, the plans and seeds, and the outputs of the analysis.
Serrano, Benjamin, Dinamarca, Alejandro, Ibacache, Claudia, Stehlik, Milan, Olivares, Rodrigo
This release provides an integrated Python workflow for predicting glass-forming ability and designing bulk metallic glass compositions under bounded compositional drift. The workflow combines machine learning, conformal calibration, composition-neighborhood sampling, and Bayesian optimization in a single script. The target property is the critical casting diameter, Dmax. Candidate compositions are ranked using a calibrated, drift-aware lower-bound estimate that accounts for changes in model predictions under mass-conserving composition perturbations. The outputs are calibrated, model-based design estimates whose interpretation depends on the
Anurag Bajpai
Experiment-Predictium v2.3.0-VM-RR Release date: 19 October 2026 Overview This release consolidates the MOF-specific development carried out during the 2026 Virtual Mobility activity "MOF-Predictium: A No-Code Machine Learning Platform for MOF Property Prediction, Benchmarking, and High-Throughput Screening Using CoRE MOF DB" within COST Action CA22147 - EU4MOFs. Main additions MOF-Predictium as an eighth model option, combining 1D convolution, multi-head self-attention, and gradient-boosted residual correction. Benchmarking with 80/20 holdout evaluation, 5-fold cross-validation, RMSE, MAE, MAPE, and R2. Advanced MOF VM Tools for multi-target
Bingöl, Alper
MATLAB execution code accompanying the manuscript "ADR-SSA: An Enhanced Sparrow Search Algorithm for Complex Optimization and Data-Driven Forecasting" . ADR-SSA enhances the Sparrow Search Algorithm with two coordinated mechanisms: dynamic reverse learning, which repositions the worst-performing individuals relative to the evolving search region, and adaptive t-distribution mutation, which perturbs elite individuals with a scale that adapts during iteration. Contents: the ADR-SSA implementation ( code/ADR_SSA.m ); CEC2022 comparison and component-ablation scripts ( Ablation.m , CEC2022_DRL_ATM_ExcelOnly.m ); constrained engineering design exp
Li, Chunbing, Wang, Jiayi, Chen, Zhenlin, Qiu, Xuerui, Zhang, Yilin, Wang, Yile
Mass spectrometry (MS) experiments generate rich acquisition metadata that are essential for reproducibility, data sharing, and quality control (QC). Because these metadata are typically stored only in vendor-specific formats, they often remain difficult to access. MetaXtract is a lightweight tool that extracts detailed parameters directly from Thermo Fisher raw files and exposes them in structured, tabular formats. By capturing sample information, LC-MS method settings, and scan-level metrics such as retention time, total ion current, and ion injection time, MetaXtract increases transparency and ensures that essential acquisition details acc
Ahmad Lutfi, Zhuo A. Chen, Lutz Fischer, Juri Rappsilber
Official Codebase Release (v1.0.1) Accompanying reproducible code, neural architectures, data pipelines, and decision-support tools for the Springer Nature Monograph : "AI-Driven Geospatial Hydrology: Machine Learning and Decision Support for Surface Water Dynamics" (2026) by Burak Can . Key Capabilities & Modules Earth Observation Pipelines: Cloud-native STAC metadata querying, COG windowed streaming, and 6S radiative transfer normalization (Chapters 1–2). Deep Vision for Water Extraction: Boundary-Aware ResUNet, Hierarchical SegFormer (MiT-B3), and multi-modal SAR-optical FCLS unmixing (Chapters 3–5). Hydrodynamic Sequence Forecasting: Diff
Can, Burak
