Shoreline Change Forecasting on Eretan Beach using Long Short Term Memory

Iryanto,; Gunawan, P. H.; Satrio, Ari; Baizal, ZK Abdurahman; Ghozali, Ahmad Lubis; Ismantohadi, Eka

doi:10.1109/icoiact55506.2022.9972113

Cited by 2 publications

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mechanism of the LSTM method in processing information can be governed in equations ( 1) -( 6) [10], [24].…”

Section: Data Modellingmentioning

confidence: 99%

“…Further, performance of the method depends on its hyperparameter. In [10], values of time step, batch size, and hidden neuron can be set as 1 -20, 2 -64, and 1 -400, respectively to get good performance.…”

Section: Data Modellingmentioning

confidence: 99%

“…Note that, study on shoreline change in Eretan beach was previously studied using LSTM in [10] and random forest in [11]. It is mentioned in [10] that optimizer functions such as RMSProp, Adam, Adamax, and Nadam are promising optimizer to increase performance of the LSTM method. Thus, those optimizer functions are used in this article.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Long Short Term Memory Approach for Shoreline Change Prediction on Eretan Beach

Iryanto,

Satrio,

Ghozali

et al. 2024

jitk

View full text Add to dashboard Cite

Eretan Beach is one of the beaches in Indramayu and has a reasonably severe abrasion rate from year to year. The Eretan coastline always experiences significant changes due to erosion every year. Therefore, it is necessary to study changes in the coastline at Eretan beach. This study obtained coastline data from the Google Earth engine using CoastSat, a python-based open-source toolkit, from 1992 – 2022. The open-source geographic information system software used to process the data is the Quantum Geographic Information System. This study aims to analyze the Long Short-term Memory (LSTM) algorithm in predicting shoreline changes at Eretan Beach. The eight optimizer functions in the LSTM are used with nine different scenarios to analyze the algorithm's performance. The results of this study show that RMSProp has the best performance compared to other optimizers. The RMSE and MAPE values on the RMSProp are 35.06258 and 2.2923 on the training data and 9.2457 and 1.06786 on the test data. In addition, from the predictions for the next ten years at transect point 251, it was found that there would be an increase in the coastline.

show abstract

“…Mechanism of the LSTM method in processing information can be governed in equations ( 1) -( 6) [10], [24].…”

Section: Data Modellingmentioning

confidence: 99%

Section: Data Modellingmentioning

confidence: 99%