Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Miltiadis Iatrou has a total of 54 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A Privacy-Preserving Digital Soil Mapping Framework for Integrating Private High-Spatial-Resolution Soil Data into EU Soil Monitoring Infrastructures

    by Panagiotis Tziachris, Tristano Bacchetti-De-Gregoris, Miltiadis Iatrou, Vasileios Takavakoglou, Karina Patricia Prazeres Marques, Vassilis Aschonitis

    High-spatial-resolution soil data collected on behalf of the private sector (e.g., farmers) represents a largely untapped resource for EU soil monitoring initiatives. Georeferenced soil samples also raise privacy issues since sampling locations can be linked to individual farm parcels and their respective activities. This work presents a Privacy-Preserving Digital Soil Mapping Framework (PP-DSM) that enables integration of private georeferenced soil datasets while releasing only aggregated spatial outputs, minimising risks of individual farm identification. The framework consists of three components: a secure soil data processing environment that keeps private point data under institutional control, a Quantile Regression Forest (QRF) engine that produces spatially explicit predictions and uncertainty estimates, and spatial aggregation of raster outputs to the EU LUCAS 2 × 2 km monitoring grid, releasing only anonymised polygon-level statistics based on polygons centred on the grid points. This methodology is demonstrated using a case study of a published georeferenced soil dataset of organic carbon (403 topsoil samples) from the Kastoria region, Northern Greece. Aggregated predictions preserved regional soil patterns while eliminating farm-level identifiability. Across six independently validated LUCAS polygons, QRF polygon statistics differed from independent test set means by an average difference of 0.070% SOC, consistent with the expected spatial smoothing. This study suggests that privately held soil datasets can support EU monitoring infrastructures within the existing regulatory environment, contributing to Soil Monitoring Law objectives and Carbon Removals and Carbon Farming initiatives.


  • Top Cited Papers
    Open Access

    Comparing XGBoost and Double Machine Learning for Predicting the Nitrogen Requirement of Rice

    by Miltiadis Iatrou, Spiros Mourelatos, Christos Karydas

    Estimating how crop yield responds to site-specific nitrogen (N) fertilization is essential for maximizing yield potential under variable field conditions. However, classical Machine Learning (ML) approaches applied to observational farm data primarily focus on yield prediction and often fail to recover causal N response due to confounding arising from non-random fertilizer application. In this study, we develop and evaluate a Causal Machine Learning (CML) framework to estimate heterogeneous N treatment effects under real commercial rice-farming conditions in the Axios River Plain, Greece. The proposed approach combines Double Machine Learning (DML) with remote sensing, soil, climatic, and management data to adjust for confounding and identify causal relationships between N inputs, Leaf Nitrogen Concentration (LNC), and yield. A doubly robust (DR) learner is used to estimate yield sensitivity to N at key agronomic thresholds, while a Causal Forest model leverages LNC to assess crop physiological N status. Results demonstrate that the CML-based framework outperforms conventional XGBoost predictive models in identifying field plots that are responsive to additional N. By integrating causal effect estimation with plant-status information, the proposed decision support system identifies zones where yield gains can be achieved through targeted N increases while avoiding overfertilization in non-responsive areas.


  • Open Access

    Yield Data Management in Rice Cultivation in Precision Agriculture Terms: The Greek Paradigm

    by Christos Karydas, Miltiadis Iatrou, Spiros Mourelatos

    A reliable protocol for comprehensive rice yield data management was established to overcome the heterogeneity and inconsistency inherent in using diverse data sources, measurement conditions, and units. This methodology defines systematic routines for data collection, cleansing, calibration, homogenization, analysis, and visualization. Data were collected over an eight-year period from various yield monitors across extensive cultivated areas within the Axios River Plain, Greece. The protocol relies on a Geographic Information System (GIS) to ensure high data integrity. Following calibration against in-situ weighing records, the data accuracy was confirmed as consistently high, ranging between 91.7% and 96.6% annually. Post-processing revealed a critical finding: the within-field yield variability (approx. 8.6%) was significantly lower than the inherent variability of underlying environmental factors (soil and spectral properties (33–35%)), indicating successful resource management by the farmers. Comparative analysis demonstrated that farms employing site-specific fertilization achieved significantly higher average yields (9.87 t/ha) compared to those using conventional, uniform fertilization (9.11 t/ha). The resulting calibrated yield maps are communicated via a web-based Farm Management Information System (FMIS). The established protocol has since been fully integrated into a commercial precision agriculture service, underscoring its practical efficiency and operational value.


  • Open Access

    Yield Data Management in Rice Cultivation: The Greek Paradigm

    by Christos Karydas, Miltiadis Iatrou, Spiros Mourelatos

    In this work, a protocol for rice yield data management was established, comprising phases and routines for data collection, cleansing, calibration, homogenization, filtering, analysis, and visualization. Establishment of a protocol was found necessary, considering the different data sources, conditions, regulations, methods, and units employed each time for monitoring rice yield. The data were collected from yield monitors mounted on harvesters, covering extensive cultivated areas in the Axios River Plain, Greece, over a period of eight years. The data were stored, processed, and analyzed within a geographic information system (GIS), while the resulting enhanced yield maps were communicated to the farmers through a farm management information system (FMIS) available on the web. Analysis of the yield data indicates higher yields -both in production and monetary terms- in those rice farms where site-specific fertilization was applied, compared to the farms with conventional, uniform fertilization. The resulting protocol has become a functional part of a commercial precision agriculture service.


  • Open Access

    Machine Learning vs. Langmuir: A Multioutput XGBoost Regressor Better Captures Soil Phosphorus Adsorption Dynamics

    by Miltiadis Iatrou, Aristotelis Papadopoulos

    Accurate prediction of soil phosphorus (P) adsorption capacity is essential for efficient fertilizer management and environmental protection. Traditional isotherm models, such as the Langmuir equation, have been widely used to quantify P sorption, but they do not adequately capture the nonlinear and multivariate nature of soil systems. This study evaluates the performance of a multi-output XGBoost regression model trained on laboratory-measured P adsorption data from 147 soils, representing a wide range of textures, pH levels, and CaCO3 contents. The model was developed to simultaneously predict P adsorption at five different equilibrium concentrations (1, 2, 4, 6, and 10 mg/L). SHAP analysis and causal discovery via DirectLiNGAM revealed that initial Olsen P concentration and sand content are the primary factors reducing P adsorption. The multi-output XGBoost model was compared against classical Langmuir isotherms using an extended dataset of 10,389 soil samples. The extended dataset was binned into four groups based on Olsen P concentrations and four groups based on sand content. This binning was based on the identification of these variables as highly influential by the XGBoost model, and on their demonstrated causal relationship with soil P sorption capacity through causal inference analysis. The XGBoost model outperformed the Langmuir model in capturing the effect of Olsen P and sand content, as it predicted a 12.6% drop in P adsorption in the very high Olsen P group and a 19.2% drop in the very high sand content groups, which are substantially higher than the reductions estimated by Langmuir isotherms. These results demonstrate that machine learning models, trained on well-designed experimental data, offer a superior alternative to classical isotherms for modeling P sorption dynamics.


  • Open Access

    Machine Learning vs Langmuir: A Multioutput XGBoost Regressor Better Captures Soil Phosphorus Adsorption Dynamics

    by Miltiadis Iatrou, Aristotelis Papadopoulos

    Accurate prediction of soil phosphorus (P) adsorption capacity is essential for efficient fertilizer management and environmental protection. Traditional isotherm models, such as the Langmuir equation, have been widely used to quantify P sorption, but they obviously fail to account for the nonlinear and multivariate nature of soil systems. This study evaluates the performance of a multi-output XGBoost regression model trained on laboratory-measured P adsorption data from 147 soils, representing a wide range of textures, pH levels, and CaCO₃ contents. The model was developed to simultaneously predict P adsorption at five different equilibrium concentrations (1, 2, 4, 6, and 10 mg/L). SHAP analysis and causal discovery via DirectLiNGAM revealed that initial Olsen P concentration and sand content are the primary factors reducing P adsorption. The multi-output XGBoost model was compared against classical Langmuir isotherms using an extended dataset of 10,389 soil samples. The extended dataset, comprising 10,389 rows, was binned into four groups based on Olsen P concentrations and four groups based on sand content. This binning was based on the identification of these variables as highly influential by the XGBoost model, and on their demonstrated causal relationship with soil P sorption capacity through causal inference analysis. The XGBoost model outperformed the Langmuir model in capturing the effect of Olsen P and sand content, as it predicted a 12.6% drop in P adsorption in the very high Olsen P group and a 19.2% drop in the very high sand content groups, which are substantially higher than the reductions estimated by Langmuir isotherms. These results demonstrate that machine learning models, trained on well-designed experimental data, offer a superior alternative to classical isotherms for modeling P sorption dynamics.


  • Open Access

    An Innovative Process Chain for Precision Agriculture Services

    by Christos Karydas, Miltiadis Iatrou, Spiros Mourelatos

    In this work, an innovative process chain is set up for the regular provision of fertilization consultation services to farmers for a variety of crops, within a precision agriculture framework. The central hub of this mechanism is a geographic information system (GIS), while a 5 × 5 m point grid is the information carrier. Potential data sources include soil samples, satellite imagery, meteorological parameters, yield maps, and agronomic information. Whenever big data are available per crop, decision-making is supported by machine learning systems (MLSs). All the map data are uploaded to a farm management information system (FMIS) for visualization and storage. The recipe maps are transmitted wirelessly to variable rate technologies (VRTs) for applications in the field. To a large degree, the process chain has been automated with programming at many levels. Currently, four different service modules based on the new process chain are available in the market.


  • Open Access

    A Comparative Analysis of Machine Learning and Pedotransfer Functions Under Varying Data Availability in Two Greek Regions

    by Panagiotis Tziachris, Panagiota Louka, Eirini Metaxa, Miltiadis Iatrou, Konstantinos Tsiouplakis

    The current study evaluates the performance of pedotransfer functions (PTFs) and machine learning (ML) algorithms in predicting the soil bulk density (BD) across two distinct regions in Greece—Kozani and Veroia—using both limited and extended sets of soil parameters. The results reveal significant regional differences in prediction accuracy. In the full dataset scenario, Veroia consistently exhibits superior predictive performance across all models (PDF RMSE: 0.104, ML RMSE: 0.095) compared to Kozani (PDF RMSE: 0.133, ML RMSE: 0.122). Generally, ML models outperform PTFs in terms of the RMSE and MAE in both regions with the full dataset. However, PTFs occasionally demonstrate higher R2 values (Veroia PTF R2: 0.35 vs. ML R2: 0.28), suggesting a better explanation of the overall variance despite larger errors. Notably, the effectiveness of ML appears to be affected by the availability of data. In Kozani, when restricted to basic soil properties, ML’s performance (RMSE: 0.129, R2: 0.16) becomes similar to that of PTFs (RMSE: 0.133, R2: 0.16). However, incorporating the full dataset substantially enhances ML’s predictive power (RMSE: 0.122, R2: 0.26). Conversely, in Veroia, the inclusion of more variables paradoxically results in a slight decline in ML performance (ML_min RMSE: 0.093, R2: 0.31 vs. ML RMSE: 0.095, R2: 0.28). These contrasting results emphasize the need for context-specific modeling strategies, careful feature selection, and caution against the assumption that more data or complexity inherently improves the predictive performance.

  • Open Access

    An Innovative Process Chain for Precision Agriculture Services

    by Christos Karydas, Miltiadis Iatrou, Spiros Mourelatos

    In this research, an innovative design is set up for regular provision of precision agriculture services, within the framework of fertilization consultancy. The central node of the chain is a geographic information system (GIS), while a 5x5m point grid is the information carrier. Decision-making is supported by machine learning algorithms developed with training of the collected big data. Potential data sources include soil samples, satellite data, yield maps, and agronomic information; while the produced maps are directed to a commercial farm management information system (FMIS) for visualization and storage and -in parallel- to variable-rate technologies (VRT) for the applications. In a large degree, the process chain is automated with Python programming language. The service has been tested exhaustively under true field conditions before is has entered the market.