Retrieving Vegetation Reradiation Patterns by Means of Artificial Neural Networks

Modeling and inferring the attenuation induced by vegetation barriers at 2G/3G/4G cellular bands using Artificial Neural Networks

“…The optimal number of neurons and the regularization parameter were tuned by means of the learning curves shown in Figure 9 All input features, xi , are scaled to lie within similar ranges, by subtracting their mean value, μ i , and dividing by their standard deviation, σ i , following (12). This process is called parameter scaling, and is a mandatory step when dealing with ANN [18].…”

“…In [12] authors experimented with ANNs to retrieve unknown parts of a partialmeasured re-radiation pattern of an individual shrub or tree. However, to the authors' knowledge, there are no previous published works trying to model propagation losses for vegetation barriers measured within large frequency ranges and based on ANNs.…”

Using artificial neural networks to scale and infer vegetation media phase functions

“…In [13] authors have demonstrated the ability of artificial neural networks (ANN) to retrieve discrete angle intervals of an incomplete re-radiation pattern. However, such model must be trained with a set of measurements in every desired frequency and/or dimensions, and thus, the problem of inferring a complete re-radiation function remains unsolved.…”

“…The aim of this paper is to propose the use of feedforward artificial neural networks (ANN) as reliable and accurate tools, which are perfectly capable to scale full vegetation phase functions to any new vegetation cell dimensions. This technique been selected among different methodologies such as polynomial regressions, because it has demonstrated its accuracy in [13] whereas it requires a lower number of coefficients and a smaller training set [14].…”