Ensemble Imputation Method for Forecasting Indonesia Sugar Dataset Using Machine Learning

Mas Syahdan Filsafan, Riyanarto Sarno, I. R. Bernadetta Raras, Agus Tri Haryono

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Imputing missing values in the Indonesian sugar dataset is crucial to mitigate biased predictions. In this study, we employ KNN imputation with N equal to 5 to address this issue and achieve optimal results. Multiple machine learning regression models are utilized to estimate sugar imports effectively, and a comparison is made with the deep learning method to determine the best-performing approach. After handling missing values, our research demonstrates improved accuracy and minimized errors when implementing machine and deep learning methods. Through KNN imputation, missing values are effectively handled, and the LSTM model is utilized to estimate sugar imports accurately. The LSTM model achieves a Mean Squared Error (MSE) of 0.0303 and Mean Absolute Error (MAE) of 0.1376 on the testing dataset. Additionally, KNN imputation reaches the highest average percentage value of 94.734, indicating the closest similarity to the original data.

Original languageEnglish
Title of host publication2023 1st International Conference on Advanced Engineering and Technologies, ICONNIC 2023 - Proceeding
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages13-17
Number of pages5
ISBN (Electronic)9798350306484
DOIs
Publication statusPublished - 2023
Event1st International Conference on Advanced Engineering and Technologies, ICONNIC 2023 - Kediri, Indonesia
Duration: 14 Oct 2023 → …

Publication series

Name2023 1st International Conference on Advanced Engineering and Technologies, ICONNIC 2023 - Proceeding

Conference

Conference1st International Conference on Advanced Engineering and Technologies, ICONNIC 2023
Country/TerritoryIndonesia
CityKediri
Period14/10/23 → …

Keywords

  • Deep learning
  • Forecasting
  • Machine Learning
  • Sugar Import

Fingerprint

Dive into the research topics of 'Ensemble Imputation Method for Forecasting Indonesia Sugar Dataset Using Machine Learning'. Together they form a unique fingerprint.

Cite this