MULTIVARIATE MULTISTEP TIME SERIES FORECASTING DENGAN ALGORITMA CNN-LSTM (STUDI KASUS: DATA CUACA DI STASIUN BMKG)

Yogi Anggara, NIM.: 19106010011 (2023) MULTIVARIATE MULTISTEP TIME SERIES FORECASTING DENGAN ALGORITMA CNN-LSTM (STUDI KASUS: DATA CUACA DI STASIUN BMKG). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (MULTIVARIATE MULTISTEP TIME SERIES FORECASTING DENGAN ALGORITMA CNN-LSTM (STUDI KASUS: DATA CUACA DI STASIUN BMKG)) 19106010011_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (3MB) | Preview
	Text (MULTIVARIATE MULTISTEP TIME SERIES FORECASTING DENGAN ALGORITMA CNN-LSTM (STUDI KASUS: DATA CUACA DI STASIUN BMKG)) 19106010011_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (6MB) | Request a copy

Abstract

Time series analysis is the analysis of a collection of data over a certain period of time in the past to understand and predict future conditions. Generally, classic models on time series are designed to forecast only one step ahead. Neural networks have the ability to learn models in general, so they can make forecasts several steps ahead for several time series data simultaneously. One of the neural network algorithms that can be used for multivariate multistep time series models is CNN-LSTM. CNN-LSTM (Convolutional Neural Network - Long Short Term Memory) is a neural network algorithm that has the ability to extract features automatically and learn patterns in sequential data without ignoring the order. In this study, CNN-LSTM was used to build a weather factor forecasting model at the BMKG station (Meteorology, Climatology and Geophysics) Sleman Regency from January 2016 to May 2022. The data is splited into training data, validation data and test data. The MAPE value of the test data for all weather factors is between 0 and 30 percent, which indicates that the forecast is quite feasible. Based on the SHAP (SHapley Additive exPlanations) value, information is obtained that the most influential weather factor for the model in the forecasting process is air humidity.

Share this knowledge with your friends :

Actions (login required)

View Item