Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
Khairunnisa, R. A Galuh
Haris, Abdul
Siregar, Riki Ruli A.
Subject
Teknik Informatika
Datestamp
2023-05-31 04:33:57
Abstract :
The salt-making process on Madura Island is produced in a traditional way that relies on
sunlight and depends on the climate to evaporate seawater that is pumped in the pegaraman
land. The phenomenon of erratic climate change on Madura Island is one of the success factors
of the salt production target that fluctuates every year. An increase in rainfall and the frequency
of rain events during the production season can disrupt and even cause a decrease in salt
productivity on Madura Island. This research is proposed with the aim of predicting weather
forecasts in Sumenep, Madura using the Long Short-Term Memory (LSTM) method. Data is
obtained from the BMKG official website from 2017-2022 at the Trunojoyo Meteorological
Station using variables of temperature, humidity, wind speed, rainfall and length of sunshine in
Sumenep. LSTM model testing is seen based on the number of hidden layers, batch size, number
of epochs, accuracy value, Mean Squared Error (MSE) and Root Mean Squared Error (RMSE).
From the tests that have been carried out, the best results with a high accuracy level of 81.33%
are found in the number of hidden layers 128, batch size 64, and epoch 100. With the lowest
MSE value of 39.62 and RMSE of 6.29. Based on testing, it also produces minimum rainfall or
the dry season range occurs from May to October. Meanwhile, the peak of the dry season occurs
in August.