Evaluating LSTM Performance on Multivariate Time Series with One-Class SVM Outlier Detection

Ragita Anillya Putri; Sugiyarto Surono; Aris Thobirin

doi:10.30595/juita.v13i2.26160

Authors

Ragita Anillya Putri University of Ahmad Dahlan Yogyakarta
Sugiyarto Surono University of Ahmad Dahlan Yogyakarta
Aris Thobirin University of Ahmad Dahlan Yogyakarta

DOI:

https://doi.org/10.30595/juita.v13i2.26160

Keywords:

OCSVM, LSTM, multivariate time series, weekly sales forecasting.

Abstract

Weekly sales forecasting plays a crucial role in retail business planning and inventory management.This study evaluates the prediction performance of a Long Short-Term Memory (LSTM) model for weekly sales forecasting after data preprocessing using standardization and outlier detection with One-Class Support Vector Machine (OCSVM) method. The independent variables used include temperature, fuel price, holidays, Consumer Price Index (CPI), and unemployment rate, with weekly sales as the target variable. The dataset is preprocessed using StandardScaler and OCSVM to detect and remove outliers before model training. The evaluation shows that the LSTM model on the clean data achieves an MSE of 0.03, an RMSE of 0.18, and an MAE of 0.11. The LSTM model demonstrates good forecasting performance when trained on cleaned data without outliers. This study provides practical insights into applying data preprocessing with OCSVM to improve the consistency of prediction models in retail time series analysis.

References

[1] S. S. W. Fatima and A. Rahimi, “A Review of Time-Series Forecasting Algorithms for Industrial Manufacturing Systems,” Jun. 01, 2024, Multidisciplinary Digital Publishing Institute (MDPI). doi: 10.3390/machines12060380.

[2] “The Analysis of Time Series_ An Introduction with R”.

[3] P. R. Fitrayana and D. R. S. Saputro, “Algoritme Clustering Large Application (CLARA) untuk Menangani Data Outlier,” PRISMA, Prosiding Seminar Nasional Matematika, vol. 5, pp. 721–725, 2022, [Online]. Available: https://journal.unnes.ac.id/sju/index.php/prisma/

[4] J. L. Serras, S. Vinga, and A. M. Carvalho, “Outlier detection for multivariate time series using dynamic bayesian networks,” Applied Sciences (Switzerland), vol. 11, no. 4, pp. 1–22, Feb. 2021, doi: 10.3390/app11041955.

[5] L. Melissa Yapanto, S. Muzfirah, Mp. M. Neny Rasnyanti Aras, and Ms. Nurhikmah Sibua, “ANALISIS DATA STATISTIK METODE DAN TEKNIK (STATISTICAL DATA ANALYSIS: METHODS AND TECHNIQUES).”

[6] KST 2020 : 2020-12th International Conference on Knowledge and Smart Technology : Jan 29 - Feb 1 @Amari Pattaya, Thailand. IEEE, 2020.

[7] N. Seliya, A. Abdollah Zadeh, and T. M. Khoshgoftaar, “A literature review on one-class classification and its potential applications in big data,” J Big Data, vol. 8, no. 1, Dec. 2021, doi: 10.1186/s40537-021-00514-x.

[8] M. R. Ayyagari, “Classification of Imbalanced Datasets using One-Class SVM, k-Nearest Neighbors and CART Algorithm,” 2020. [Online]. Available: www.ijacsa.thesai.org

[9] M. J. Zaki Wagner Meira Jr, “Data Mining and Analysis: Fundamental Concepts and Algorithms.”

[10] A. De Martino and K. Diki, “Gaussian RBF-kernels via Fock spaces: quaternionic and several complex variables settings,” Aug. 2023, [Online]. Available: http://arxiv.org/abs/2308.14110

[11] E. P. Wahyuddin, R. E. Caraka, R. Kurniawan, W. Caesarendra, P. U. Gio, and B. Pardamean, “Improved LSTM hyperparameters alongside sentiment walk-forward validation for time series prediction,” Journal of Open Innovation: Technology, Market, and Complexity, vol. 11, no. 1, Mar. 2025, doi: 10.1016/j.joitmc.2024.100458.

[12] X. Song et al., “Time-series well performance prediction based on Long Short-Term Memory (LSTM) neural network model,” J Pet Sci Eng, vol. 186, Mar. 2020, doi: 10.1016/j.petrol.2019.106682.

[13] M. Pala and E. Sefer, “NFT price and sales characteristics prediction by transfer learning of visual attributes,” Journal of Finance and Data Science, vol. 10, Dec. 2024, doi: 10.1016/j.jfds.2024.100148.

[14] A. Tawakuli, B. Havers, V. Gulisano, D. Kaiser, and T. Engel, “Survey:Time-series data preprocessing: A survey and an empirical analysis,” Journal of Engineering Research (Kuwait), 2024, doi: 10.1016/j.jer.2024.02.018.

[15] A. Razaki and Y. Herry Chrisnanto, “PENANGANAN OUTLIER PADA METODE ALGORITMA K-NEAREST NEIGHBORS (KNN) DENGAN METODE KERNEL DENSITY ESTIMATION PADA KASUS PENYAKIT DIABETES HANDLING OUTLIERS IN THE K-NEAREST NEIGHBORS (KNN) ALGORITHM USING KERNEL DENSITY ESTIMATION IN DIABETES CASES,” Journal of Information Technology and Computer Science (INTECOMS), vol. 7, no. 4, 2024.

[16] B. Hakim, “Analisa Sentimen Data Text Preprocessing Pada Data Mining Dengan Menggunakan Machine Learning,” JBASE - Journal of Business and Audit Information Systems, vol. 4, no. 2, Aug. 2021, doi: 10.30813/jbase.v4i2.3000.

[17] S. S. Todkar, V. Baltazart, A. Ihamouten, X. Dérobert, and D. Guilbert, “One-class SVM based outlier detection strategy to detect thin interlayer debondings within pavement structures using Ground Penetrating Radar data,” J Appl Geophy, vol. 192, Sep. 2021, doi: 10.1016/j.jappgeo.2021.104392.

[18] M. J. Zaki Wagner Meira Jr, “Data Mining and Analysis: Fundamental Concepts and Algorithms.”

[19] K. E. ArunKumar, D. V. Kalaga, C. Mohan Sai Kumar, M. Kawaji, and T. M. Brenza, “Comparative analysis of Gated Recurrent Units (GRU), long Short-Term memory (LSTM) cells, autoregressive Integrated moving average (ARIMA), seasonal autoregressive Integrated moving average (SARIMA) for forecasting COVID-19 trends,” Alexandria Engineering Journal, vol. 61, no. 10, pp. 7585–7603, Oct. 2022, doi: 10.1016/j.aej.2022.01.011.