Indonesian Jourual of Electrical Engineering and Computer Science Vol. 37, No. 1, Jemuary 2025, pp. 540~550 ISSN: 2502-4752, DOL: 10.11591ijeees. il-pp540-550 a 540 Classification of weather conditions based on automatic weather station data using a multi-layer perceptron neural network Muhammad Aristo Indrajaya, Tan Suryani Sollu, Mery Subito, Yuli Asmi Rahmay Erwin Ardias Saputra Depurtint of Elecusal Engineering, Faculty of Engineering, Unveraty of Tadulako, Pau, does Article Info ABSTRACT Article history: Weather is one of the important elements that greatly determines Inman . activities, especially those related ta economic factors. Therefore Received Tun 30, 2024 inderstanding weather conditions using weather parameters asa reference is Revised Sep 11, 2024 important for Inman life, sos method is needed to classify weather ‘Accepted Sep 29, 2024 according to its category so thatthe information produced can be used for ‘various nceds. Deteimining weather conditions in an area will not nan well without a reliable method that can analyze existing weather parameters Keywords: ‘Therefore, in this study, the weather condition classification process was curred out using the multilayer perceptron algorithm, a type of neural Accuracy y network (NN) algorithm. All data analyzed were wenther parameter dat Altair AI studio collected by mini weather stations placed on land, The weather parameters Classification asd te sapere ania, aia nd cee, eel stares ‘Multi-layer perceptron chil, daily rainfall, solar radiation, an UV index. This study was conducted ‘Weather sn Pau ity. Central Sulawesi Province, Indonesia. The clasiication process casied out by the multilayer perceptron algorithm was earied out on the Altair AI Studio application and produced an accuracy value of 93.87%, fecal of 92.33% and precision of 91 29% Ths ts an open access article under the CC BY-SA ticense. EOS Corresponding Author: Mery Suiito Department of Electrical Engineering, Faculty of Engineering, University of Tadulako Sockarno-Hatta Road Kilometer 9, Palu 941 12, Central Sulawesi, Indonesia Email: [email protected] 1. INTRODUCTION Weather is a dynamic aspect of the Earth's atmosphere that changes frequently. Temperature, air ‘pressure, humidity, light intensity, and rainfall all have an impact on the weather ina given location. Weather conditions greatly determine whether or not community activities run smoothly [1]. In the field of ‘wansportation, especially aviation and sea transportation, weather conditions are critical in terms of safety [2]. Many accidents, especially in the aviation sector, are indirectly caused by the influence of weather, so weather prediction is significant in supporting the transportation sector, especially in sea and air ‘transportation, In the agricultural and plantation sectors, the role of weather in determining agricultural production results is huge [3], [4]. Whether that is in good condition will encourage maxinnum agricultural production and provide large economic benefits for farmers [S], [6]. On the other hand, extreme weather such as long droughts and extreme rains will cause crop failures which have implications for a significant reduction in profits for farmers and under certain conditions will cause scarcity of several agricultural commodities that are needed by the community. Therefore, the ability to predict weather conditions is ‘essential in anticipating the influence of weather on the economic activities of a region [7], [8] ‘Journal homepage: hitp://ijeecs.iaescore.comIndonesian J Elec Eng & Comp Sei ISSN: 2502-4752 oa Data mining is a process that comprises data collecting and the application of past data to discover ‘an order, pattem, oF relationship in massive data 9]. Data mining involves four disciplines: statistics, ‘visualization, databases, and machine leaming. Machine learning is a subset of artificial intelligence that