Impedance-Based Assessment of Ipomoea Batatas Leaf Water Status Using Hybrid Mutual Dependence and Variance Inflation Factor Regression Analysis

This study developed an accessible electrical impedance measurement system and optimized a regression approach for rapid, nondestructive assessment of water status in sweet potato (Ipomoea Batatas) leaves. A hybrid mutual dependence-variance inflation factor (MDVIF) regression model was developed to address multicollinearity and select features when correlating leaf impedance measurements across 10–100 kHz with relative water content (RWC) measured under controlled water stress conditions. The MDVIF model achieved an R2 of 0.97, identifying critical impedance frequencies correlated with RWC. This MDVIF approach demonstrated superior predictive performance compared with wrapper feature selection, with a lower root mean square error (RMSE) of 0.033, accurately capturing the complex impedance-water relationship. The low-cost, portable system integrated with the optimized regression algorithm shows potential for practical field-based monitoring of crop water status, providing a rapid, noninvasive evaluation of plant water status.