Analysis of potential hazards at the sea with artificial neural network and accident prevention with SOS smart system innovation

• Main Authors: Harumy, T. H. F., Ginting, Dewi Sartika Br, Handrizal, Albana, M. F., Jamesie, A. B., Patrecella, R. P.
• Format: Conference Proceeding
• Language: English
• Subjects: Accident prevention; Artificial neural networks; Data collection; Decision trees; Economic factors; First aid; Health services; Hierarchies; Innovations; Mapping; Multilayer perceptrons; Neural networks; Public health; Social factors; User interfaces
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0200202

Currently, there are still many incidents that happen to fishermen while at sea, such as being bitten by venomous sea snakes, storms that cause death without first aid due to lack of education and innovation of first aid systems that can help the fishermen. This is also caused by several conditions, namely social factors, environmental factors, economic factors and habits of coastal communities. So that innovation is needed by analyzing in depth the design of intelligent systems that can be used by the community. Innovation of SOS smart systems and mapping of health service location searches based on Artificial Neural Network which will be applied to the relevant environment based on the model developed in an effort to improve the quality of public health services. coast. The formulation of the problem is how to analyze potential incidents that occur at sea and build a smart social system innovation & mapping of finding the nearest health service location based on an artificial intelligent neural network that can be used to improve the quality of health services for coastal communities. The stages of the research method are Attribute/Variable Data Collection and Development of a Literature Study Model (Secondary Data), Interviews with Respondents (Coastal Communities), Field Observation, Smart System UI/UX Design, Application of the Neural Network Method on Smart System SOS and looking for locations nearest health service and Analysis and Testing for Accuracy and Compatibility of Models and Features on Smart System SOS. The data sample used is 50 data. From the results of the study, it can be stated that the Potential Incidents at Sea are 19 or 38% who are categorized as moderate, 16 or 32% of people are categorized as low and 15 people or 30% are categorized as exposed at sea. Furthermore, the best method in classifying Potential Incidents at Sea is Neural Network Multilayer Perceptron by 94%, then Naïve Bayes by 60%% and Decision tree by 70%. So that the neural Network can be implemented to be implemented in SOS smart systems.