Article: A novel framework for forecasting time series data based on fuzzy logic and variants of hidden Markov model Journal: International Journal of Business Intelligence and Data Mining (IJBIDM) 2021 Vol.19 No.3 pp.304 - 318 Abstract: The traditional time series forecasting methods such as naive, smoothing model and moving average model assumes that time series is stationary and could not handle linguistic terms. To provide solution to this problem, fuzzy time series (FTS) forecasting methods are being considered in this research work. The objective of this research is to improve the accuracy by introducing a new partitioning method called relative differences-(RD) based interval method. This research work implements the variants of RD-based hidden Markov models (HMM) such as classic HMM, stochastic HMM, Laplace stochastic smoothing HMM and probabilistic smoothing HMM (PsHMM) for forecasting time series data. In the proposed work, the performances of the above models were tested with Australian electricity market dataset and Tamil Nadu weather dataset. The results show that the performance of the proposed model namely the relative differences-based probability smoothing hidden Markov model (RD-PsHMM) performs much better in terms of precision than other existing models. Inderscience Publishers - linking academia, business and industry through research

Title: A novel framework for forecasting time series data based on fuzzy logic and variants of hidden Markov model

Authors: S. Sridevi; S. Parthasarathy; T. Chandrakumar; S. Rajaram

Addresses: Department of Information Technology, Thiagarajar College of Engineering, Madurai, Tamil Nadu, India ' Department of Computer Applications, Thiagarajar College of Engineering, Madurai, Tamil Nadu, India ' Department of Computer Applications, Thiagarajar College of Engineering, Madurai, Tamil Nadu, India ' Department of Electronics and Communications Engineering, Thiagarajar College of Engineering, Madurai, Tamil Nadu, India

Abstract: The traditional time series forecasting methods such as naive, smoothing model and moving average model assumes that time series is stationary and could not handle linguistic terms. To provide solution to this problem, fuzzy time series (FTS) forecasting methods are being considered in this research work. The objective of this research is to improve the accuracy by introducing a new partitioning method called relative differences-(RD) based interval method. This research work implements the variants of RD-based hidden Markov models (HMM) such as classic HMM, stochastic HMM, Laplace stochastic smoothing HMM and probabilistic smoothing HMM (PsHMM) for forecasting time series data. In the proposed work, the performances of the above models were tested with Australian electricity market dataset and Tamil Nadu weather dataset. The results show that the performance of the proposed model namely the relative differences-based probability smoothing hidden Markov model (RD-PsHMM) performs much better in terms of precision than other existing models.

Keywords: forecasting; time series; fuzzy; time variant model; Markov model.

DOI: 10.1504/IJBIDM.2021.118193

International Journal of Business Intelligence and Data Mining, 2021 Vol.19 No.3, pp.304 - 318

Received: 20 Mar 2019
Accepted: 27 Aug 2019
Published online: 15 Oct 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: A novel framework for forecasting time series data based on fuzzy logic and variants of hidden Markov model

Keep up-to-date