Skripsi

MACHINE LEARNING UNTUK PREDIKSI TEMPORAL PERUBAHAN IKLIM: STUDI KASUS DATA STASIUN KLIMATOLOGI PALEMBANG

---

This study aims to predict climate parameters in Palembang City using a Multivariate Long Short-Term Memory approach based on climatological data from 1981 to 2024. The model was developed to predict maximum temperature, average relative humidity, average wind speed, maximum wind direction, duration of sunshine, and rainfall. The evaluation results show that the Long Short-Term Memory model is capable of providing fairly good predictions for maximum temperature (coefficient of determination R² = 0.62), average relative humidity (R² = 0.60), average wind speed (R² = 0.61), and wind direction at maximum speed (R² = 0.76). The variable of sunshine duration was predicted with moderate accuracy (R² = 0.47), while rainfall showed poor performance (R² = 0.29), confirming the model's limitations in projecting the stochastic phenomenon of tropical rainfall. Overall, this study provides an overview of the predictability of climate parameters in Palembang and indications of a long-term trend of increasing temperatures. The applied contribution is realized through the development of an interactive dashboard based on Streamlit that can be used as an initial instrument in climate change adaptation studies at the local level. Keywords: Long Short-Term Memory, climate prediction, temporal, Palembang, climate change

---

Availability

Detail Information

Series Title

-

Call Number

T1912382025

Publisher

Indralaya : Prodi Pendidikan Fisika, Fakultas Keguruan dan Ilmu Pendidikan Universitas Sriwijaya., 2025

Collation

xvii, 103 hlm.; ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

551.607

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Meteorologi
Prodi Pendidikan Fisika

Specific Detail Info

-

Statement of Responsibility

TUTI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment