Aditya Wisnugraha Sugiyarto, Achmad Ramadhanna’il Rasjava
Flooding is one of the natural disasters that poses a serious threat to people who live near watersheds and coastal areas. Floods may affect various sectors of socioeconomic life of a society, such as sectors of the economy, agriculture, and education. This led to the need for an accurate method for predicting flood-prone areas so the public and the government can prevent and minimize the negative impacts and able to assist the recovery process in the sectors affected by the floods more optimal. In the Industrial Revolution 4.0 era, science and technology develop very quickly, one example is the development of Artificial Intelligence (AI) where the system adopts the way of human thinking, such as planning, learning, reasoning, and self-correction which is then manifested in a mathematical form so it can be applied for solving real problems. Therefore, this study used the AI system in the process of prediction and mapping of disaster-prone areas. The data used in this study are rainfall, land altitude, watersheds, river depth, and distance of the settlement to the seashore or river. These data were then processed using the method of deep learning for prediction function and Fuzzy C-Means (FCM) for mapping function which is a technique in AI systems. The final results of this study were obtained three criteria for disaster vulnerability, namely low, medium, and high, which can be used to predict the mapping of flood-prone areas until the coming years in the Bantul Regency area. © 2021, Springer Nature Singapore Pte Ltd.
Department of Mathematics, Universitas Negeri Yogyakarta, Yogyakarta, Indonesia; Department of Chemistry, Universitas Negeri Yogyakarta, Yogyakarta, Indonesia
