{Roshni Kamath Caroline Arnold Frauke Albrecht Rıdvan Salih Kuzu}

Abstract

The AI4EO HYPERVIEW challenge seeks machine learningmethods that predict agriculturally relevant soil parameters(K, Mg, P2O5, pH) from airborne hyperspectral images.We present a hybrid model fusing Random Forest and K-nearest neighbor regressors that exploit the average spectralreflectance, as well as derived features such as gradients,wavelet coefficients, and Fourier transforms. The solution iscomputationally lightweight and improves upon the challengebaseline by 21.9%, with the first place on the public leader-board. In addition, we discuss neural network architecturesand potential future improvements.