比较机器学习方法来估计土壤和液压导电性
Ali Akbar Moosavi1, Mohammad Amin Nematollahi2, Mohammad Omidifard1
1Faculty of Agriculture, Department of Soil Science and Engineering, Shiraz University, Shiraz, IR Iran.
PloS one
|November 14, 2024
概括
机器学习模型,特别是粒子群优化神经网络 (PSO-NN),使用易于测量的土壤特性准确预测土壤液压导电性 (Kfs). 这些先进的方法为水文建模的传统实验提供了更有效的替代方案.
科学领域:
- 土壤科学 土壤科学
- 水文学的水文学
- 机器学习 机器学习
背景情况:
- 准确地描述近 (场) 和液压导电性 (Kfs) 对水文建模至关重要.
- 对于Kfs的实验室和实地实验是耗时和劳动密集的.
- 脚转移函数 (PTF) 是用于根据易于测量的土壤属性预测Kfs的统计工具.
研究的目的:
- 评估各种机器学习方法的有效性,包括人工神经网络 (ANN),用于预测Kfs.
- 为了比较不同ANN和传统回归模型在预测Kfs.的性能.
主要方法:
- 测量了100个样本的土壤物理化学特性 (例如,散密度,含水量,聚合物大小,pH,EC,CCE).
- 开发了人工神经网络模型,包括辐射基函数 (RBFNNs),多层感知器 (MLPNNs),遗传算法神经网络 (GA-NNs) 和粒子群优化神经网络 (PSO-NNs).
- 使用统计指数评估模型准确性,并与多重线性回归 (MLR) 模型进行比较.
主要成果:
- 粒子集群优化神经网络 (PSO-NN) 在预测Kfs方面表现出最高的准确性,根平均平方误差 (RMSE) 和平均绝对百分比误差 (MAPE) 最低,相关系数 (R) 最高.
- 这些预测模型的排名是:PSO-NNs (R=0.958),GA-NNs (R=0.949),MLPNNs (R=0.933),RBFNNs (R=0.926) 和MLR (R=0.675).
- 所有的神经网络模型,特别是PSO-NNs,都被证明对Kfs预测有效.
结论:
- 机器学习模型,特别是PSO-NNs,对于预测土壤液压导电性 (Kfs) 是非常有效的.
- 这些模型为水文研究中Kfs估计的传统方法提供了强大而高效的替代方案.
- 建议进行进一步的研究,以评估这些模型在不同土壤条件和地理位置的更广泛适用性和潜在不确定性.
相关概念视频
01:23Measurement of Air Content in Concrete
Air content measurement in concrete is critical for ensuring structural integrity and durability of concrete structures, especially in environments prone to severe weather conditions. Accurate air content analysis optimizes concrete's resistance to freeze-thaw cycles and enhances its workability and strength. Several methods are standardized under ASTM guidelines to measure the air content in fresh concrete, each suitable for different concrete types and conditions.
The pressure method,...
The pressure method,...
01:29Application of Linearization and Approximation
A drone flying through complex terrain often relies on more than one sensing method to estimate small changes in altitude. Along with direct measurements, air pressure provides a useful indirect indicator of vertical movement. Atmospheric pressure decreases as altitude increases, and this relationship is commonly described using an exponential model. Although accurate, converting pressure measurements into altitude values requires calculations that are too complex to perform repeatedly during...