Investigation of Long Short-Term Memory Artificial Neural Networks as Estimators of Nitrate Concentrations in Soil from Measured Electrical Impedance Spectra

Monitoring the nitrate concentration in soil is crucial to guide the use of nitrate-based fertilizers. This study presents an investigation of long short-term memory (LSTM) recurrent artificial neural networks with regard to their suitability to extract nitrate concentrations from electrical impedance spectra of soil samples. Based on measured impedance spectra and physical properties of various synthetic sandy soils, the importance of different features for model training was investigated first. Both Random Forests and LSTM were tested as feature selection methods. Then numerous LSTM networks were trained to predict the nitrate concentration in sandy soils. The resulting regression models showed coefficients of determination between true and predicted nitrate concentrations as high as 0.95.

[1]  S. Duan,et al.  Short-Term Forecasting of Photovoltaic Power Generation Based on Feature Selection and Bias Compensation–LSTM Network , 2021, Energies.

[2]  Xu Wang,et al.  Short‐term building load forecast based on a data‐mining feature selection and LSTM‐RNN method , 2020, IEEJ Transactions on Electrical and Electronic Engineering.

[3]  Cody B. Hyndman,et al.  NEU: A Meta-Algorithm for Universal UAP-Invariant Feature Representation , 2018, J. Mach. Learn. Res..

[4]  Jürgen Schmidhuber,et al.  LSTM: A Search Space Odyssey , 2015, IEEE Transactions on Neural Networks and Learning Systems.

[5]  Rui Jiang,et al.  A random forest approach to the detection of epistatic interactions in case-control studies , 2009, BMC Bioinformatics.

[6]  Chong Jin Ong,et al.  A Feature Selection Method for Multilevel Mental Fatigue EEG Classification , 2007, IEEE Transactions on Biomedical Engineering.

[7]  Bjoern H Menze,et al.  Multivariate feature selection and hierarchical classification for infrared spectroscopy: serum-based detection of bovine spongiform encephalopathy , 2007, Analytical and bioanalytical chemistry.

[8]  C. Furlanello,et al.  Recursive feature elimination with random forest for PTR-MS analysis of agroindustrial products , 2006 .

[9]  Pierre Geurts,et al.  Proteomic mass spectra classification using decision tree based ensemble methods , 2005, Bioinform..

[10]  Jürgen Schmidhuber,et al.  LSTM can Solve Hard Long Time Lag Problems , 1996, NIPS.

[11]  L. Breiman.  Random Forests , 2001, Machine Learning.