ESTIMATION MODEL OF COVID-19 DAILY CASES USING LSTM (LONG SHORT-TERM MEMORY) METHOD AND 1 DIMENSIONAL CONVOLUTIONAL NEURAL NETWORK METHOD

• Main Author: SYARIEF SYAFIE, ACHMAD
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/62562
• Institution: Institut Teknologi Bandung

Coronavirus Disease 2019 or known as COVID-19 has spread rapidly throughout the world including Indonesia and has infected millions of people. Various plans have been prepared in response to the daily cases of Covid-19 in Indonesia, including implementing social distancing protocols, PPKM, vaccination, etc. Using the Indonesian movement range index data from Facebook, vaccination data and variants of Covid-19 in Indonesia obtained from our world in data, it is hoped that a good estimation model for the daily cases of Covid-19 in Indonesia can be made. The author uses 2 methods, Long Short-Term Memory (LSTM) and 1 Dimensional Convolutional Neural Network (1D-CNN) to create 5 models, including Single step Univariate LSTM, Multistep Univariate LSTM, Multistep Multivariate LSTM without additional Covid-19 variant factors, Multistep Multivariate 1D-CNN and Multistep Multivariate LSTM with additional Covid-19 Variant factor. The five models created can estimate the daily cases of Covid-19 in Indonesia, but the single step univariate and multistep univariate models which are relatively simpler, show better results when compared to the multivariate LSTM and 1D-CNN models.