Data and Code for: Machine learning-enabled implantable plant biomarker sensor for early detection and classification of acid and salt stress

This repository contains the dataset and custom machine learning code for the Machine Learning-Enabled Implantable Plant Biomarker Sensor. The framework enables the early and high-precision diagnosis of plant stress by directly monitoring internal physiological markers, detecting anomalies up to 48 hours before visible phenotypic symptoms appear. DatasetThe dataset consists of continuous in vivo electrochemical measurements from 60 hydroponic lettuce (Lactuca sativa) plants. Treatments (6 Groups): Control, Acid stress (pH 3.0, 2.0), Salt stress (300 mM, 500 mM NaCl), and Combined stress (pH 2.0 + 300 mM NaCl). Data Collection: 8-hour core monitoring window post-stress application at a 0.1 Hz sampling rate. Preprocessing: Raw data was augmented using baseline correction and a 100-second sliding window, yielding over 160,000 independent data points for robust model training.