Optimization of Flood Prediction using SVM Algorithm to determine Flood Prone Areas

Flooding is one thing that can slow down the economic pace in the affected area. Bandung is called the city of flowers and the city of fashion because the nickname makes Bandung a city with a variety of fashions growing in multiple places as a starting point for the buying and selling process. Not only did Bandung spawn fashions that became hits every year, but it also had many Meccas of traditional food preparation that were extraordinarily unique and interesting. Creating a flood-prone area model can make it easier to provide information for communities in Bandung Prefecture that belong to flood-prone and non-flood-prone areas. The SVM algorithm is a technique that can be used in the case of classification and regression, which is very popular lately. SVM is in a class with Artificial Neural Networks (ANN) in terms of features and conditions of problems that can be solved, and to be able to increase its accuracy it uses what can be optimized with PSO (Particle Swarm Optimization), where the test data is used BNPB official website data, BPS Bandung District and BMKG processed. The accuracy rate generated by using the SVM algorithm is 85.71% and the generated AUC is 0.841, while the accuracy rate generated by using the PSM-based SVM algorithm is 97.62%. and AUC produced at 1,000.