Enhanced interpretation of magnetic survey data from archaeological sites using artificial neural networks

Abstract

The use of magnetic surveys for archaeological prospecting is a well-established and versatile technique, and a wide range of data processing routines are often applied to further enhance acquired data or derive source parameters. Of particular interest in this respect is the application of artificial neural networks (ANNs) to predict source parameters such as the burial depths of detected features of interest. Within this study, ANNs based upon a multilayer perceptron architecture are used to perform the nonlinear mapping between buried wall features detected within the magnetic data and their corresponding burial depth for surveys in the ancient city of Butrint in southern Albania, achieving a greater level of information from the survey data.Suitable network training examples and test data were generated using forward models based upon ground-truth observations. The training procedure adopts a supervised learning routine that is optimized using a conjugate gradient method, while the learning algorithm also prunes network elements to prevent overregularization by reducing model complexity. Data processing was further enhanced by introducing rotational invariance using Zernike moments and by utilizing the combined output of a number, or committee, of networks. When applied to a section of survey data from Butrint, the ANN routine successfully predicted the burial depth of a number of detected wall features, with an rms error on the order of 0.20 m, and provided a coherent map of the buried building foundations. The neural network approach offered advantages in terms of efficiency and flexibility over more conventional data-inversion techniques within the context of the study, giving fast solutions for large, complex data sets while having high noise tolerance

Interpretation of magnetic anomalies due to dipping dikes using neural network inversion

“…In inverse theory, a geometrical model is chosen with initial estimates of the causative target parameters, and then the process is iteratively progressed until a satisfactory fit between the observed and the calculated anomalies is obtained (Dondurur and Pamukçu 2003). The inversion is implemented by constructing a correspondence magnetic model of the subsurface target, having a number of modifiable parameters optimized, to obtain a satisfactory estimation of the modeled data (Bescoby et al 2006). Several numerical methods have been developed such as the singular value decomposition, gradient, Gauss-Newton, Marquardt-Levenberg, and ridge regression methods have been utilized to estimate the body parameters automatically.…”

تفسيرات بعض الاجسام المغناطيسية بالاسلوب التعاكسي مستعملا الشبكات العصبية

“…Furthermore, the NNs have been successfully implemented for inverting and processing seismic data (Calderόn-Macías et al 2000;Poulton 2002); parameter estimation from well logs (Helle et al 2001); and electromagnetic, DC resistivity, and gravity geophysical data (Poulton et al 1992a, b;ElKaliouby and Poulton 1999;El-Kaliouby 2001;Spichak and Popova 2000;El-Qady and Ushijima 2001;Ucan et al 2002). The inversion of analysis of magnetic data has been implemented by Fossati et al (1992), Guo et al (1992), Ucan et al (2002), and Bescoby et al (2006).…”

“…In inversion theory, it is required to have initial values and then progress iteratively until obtaining a satisfactory fit between the observed data and the calculated ones. The inversion is often implemented by constructing a corresponding magnetic model of the subsurface causative target, including a number of modified parameters optimized to obtain a satisfactory estimation of the data (Bescoby et al 2006). Methods such as singular value decomposition, gradient, Gauss-Newton, MarquardtLevenberg, and ridge regression methods have been used to estimate the causative target parameters automatically.…”

“…However, the training data set must be designed carefully for maximum flexibility. These characteristics make it possible for NN to solve complex (large-scale) problems that are currently intractable (Bescoby et al 2006).…”

“…Generally, the inversion is implemented using a corresponding magnetic model of the causative target, containing a number of adjustable parameters optimized to estimate the data being modeled (Bescoby et al 2006).…”