ElectroML: an open-source web platform for machine learning-based analyte concentration prediction in electrochemical sensing
SOFTWAREX, vol.34, 2026 (SCI-Expanded, Scopus)
- Publication Type: Article / Article
- Volume: 34
- Publication Date: 2026
- Doi Number: 10.1016/j.softx.2026.102640
- Journal Name: SOFTWAREX
- Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC, Directory of Open Access Journals
- Istanbul University Affiliated: No
Abstract
ElectroML is an open-source web platform for machine-learning-based analyte concentration prediction from electrochemical sensing data. The platform extracts key features from cyclic voltammetry (CV) and differential pulse voltammetry (DPV) signals and offers preprocessing tools such as outlier detection, baseline correction, and signal filtering. The platform integrates multiple machine learning (ML) algorithms with automated hyperparameter optimization and supports Leave-One-Out (LOO) and K-Fold cross-validation strategies. ElectroML enables flexible concentration prediction with bootstrap-based confidence intervals and offers customizable, publication-ready visualizations with adjustable typography, color schemes, and vector export options. Its webbased interface simplifies access to advanced electrochemical analysis and supports integration with laboratory information systems. In addition to training models on existing datasets, ElectroML can rapidly predict analyte concentrations from new sensor measurements, making it a valuable tool for environmental monitoring, medical diagnostics, and industrial quality control.