Editorial for Special Issue: “Application of Artificial Neural Networks in Geoinformatics”

1. Introduction

Recently, a need has arisen for prediction techniques that can address a variety of problems by combining methods from the rapidly developing field of machine learning with geoinformation technologies such as GIS, remote sensing, and GPS. As a result, over the last few decades, one particular machine learning technology known as artificial neural networks has been successfully applied to a wide range of fields in science and engineering. In addition, the development of computational and spatial technologies has led to the rapid growth of geoinformatics, which specializes in the analysis of spatial information. Thus, recently, artificial neural networks have been applied to geoinformatics and have produced valuable results in the fields of geoscience, environment, natural hazards, natural resources, and engineering. Hence, this special issue of the journal Applied Sciences, “Application of Artificial Neural Networks in Geoinformatics,” was successfully planned, and we here publish many papers detailing novel contributions that are of relevance to these topics.

2. Applications of Artificial Neural Networks in Geoinformatics

In total, 23 papers were submitted to this special issue, 14 of which were accepted and published, constituting a 61% acceptance rate. The papers in this special issue cover various areas related to the application of artificial neural networks to GIS, remote sensing, and GPS, which are typical tools used by geoinformation researchers. These papers addressed problems such as the detection, assessment, and prediction of landslides, volcanos, forest, ozone, oil spills, buildings, ships, habitat, and traffic.

Four papers used GIS tools with artificial neural networks. The first and second papers, authored by Lee, S., Lee, M., Jung, H. [1] and Oh, H., Lee, S. [2], applied GIS and various machine learning algorithms such as artificial neural networks, support vector machines, and boosted tress to map landslide susceptibility. The third paper, authored by Lee, S., Lee, S., Song, W., Lee, M. [3], applied GIS with artificial neural networks to map potential marten and leopard habitats. The fourth paper, authored by Shah, S., Brijs, T., Ahmad, N., Pirdavani, A., Shen, Y., Basheer, M. [4], used data envelopment analysis in GIS with artificial neural networks to evaluate risks related to road safety.

Seven papers studied the applications of artificial neural networks to remote sensing. Among these, three papers used various image technologies and artificial neural networks for the detection of landslides, oil spills, and ships. Mezaal, M., Pradhan, B., Sameen, M., Mohd, S. H., Yusoff, Z. [5] used airborne laser scanning images to detect landslides. Chen, G., Li, Y., Sun, G., Zhang, Y. [6] used polarimetric synthetic aperture radar images to detect oil spills. Hwang, J., Chae, S., Kim, D., Jung, H. [7] used X-band Kompsat-5 images to detect ships. Additionally, Piscini, A., Romaniello, V., Bignami, C., Stramondo, S. [8] proposed a damage assessment method based on SAR and Sentinel-2 images, and Kadavi, P., Lee, W., Lee, C. [9] analyzed pyroclastic flow deposits using Landsat images. Kwon, S., Jung, H., Baek, W., Kim, D. [10] classified the vertical structures of forests using aerial orthophoto and Lidar images. Finally, Foody, G. [11] analyzed the impact of sample design on data validation using remote sensing data classified by feedforward neural networks, and then used a validation dataset to test the classification accuracy.

As the another geoinformation tool, GPS was used for real-time transportation mode identification with artificial neural networks by Byon, Y., Ha, J., Cho, C., Kim, T., Yeun, C. [12]. The paper authored by Sameen, M., Pradhan, B. [13] applied artificial neural networks to predict traffic accident recurrence, and Afonso, N., Pires, J. [14] applied artificial neural networks and genetic algorithms to characterize surface ozone behavior.

3. Future of Artificial Neural Networks in Geoinformatics

In this special issue, we only included papers on artificial neural networks and geoinformation technology. However, artificial neural networks are just one machine learning technique, albeit one of the most popular. Machine learning is a field of computer science that gives computers the ability to learn without being explicitly programmed. Machine learning explores the study and construction of algorithms that can learn from data and make data-driven predictions or decisions by building a model from sample inputs. There are numerous machine learning techniques, such as decision trees, support vector machines, naive Bayes classifier, clustering, inductive logic programming, and genetic algorithms. These machine learning techniques can be combined with geoinformation technologies, and further studies are required in this area.

As we enter the age of the fourth industrial revolution, artificial intelligence technologies have come to play a very important role in society. Machine learning technologies such as artificial neural networks are expected to play a key role in the fourth industrial revolution, especially in combination with geoinformation technology. However, these technologies do not come out of nowhere; they are developed by scientists. Therefore, many scientists will have to expand on the research presented in this special issue.