Time Series Forecasting Models: A Comprehensive Review
Devyani Rawat1, Vijay Singh2, Shiv Ashish Dhondiyal3, Sumeshwar Singh4
1Devyani Rawat, Department of Computer Science and Engineering, Graphic Era Deemed to be University, Dehradun (Uttarakhand), India.
2Vijay Singh, Department of Computer Science and Engineering, Graphic Era Deemed to be University, Dehradun (Uttarakhand), India.
3Shiv Ashish Dhondiyal, Department of Computer Science and Engineering, Graphic Era Deemed to be University, Dehradun (Uttarakhand), India.
4Sumeshwar Singh, Department of Computer Science and Engineering, Graphic Era Hill University, Dehradun (Uttarakhand), India.
Manuscript received on 16 June 2019 | Revised Manuscript received on 23 June 2019 | Manuscript Published on 01 July 2020 | PP: 84-86 | Volume-8 Issue-2S12 September 2019 | Retrieval Number: B10150982S1219/2020©BEIESP | DOI: 10.35940/ijrte.B1015.0982S1219
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: This comprehensive review provides an extensive overview of the existing Time Series Forecasting technique. This survey is not restricted to any single time series analysis; it provides forecasting of time series in different areas like marketing prediction, weather forecasting, technology prediction, financial forecasting etc. In this paper, we have analyzed forecasting in some areas namely, load forecasting, wind speed forecasting, prediction of energy consumption and short-term traffic flow prediction. Various models are available for prediction among them Autoregressive Integrated Moving Average model (ARIMA) is seen as a universal mechanism, these discussed forecasting areas utilizes different models that are combined with ARIMA. Hybrid models are the combination of classical models and modern methods, like ARIMA (classical method) combines with Artificial Neural Network (ANN) as well as with Support Vector Machine (SVM) (modern models). Hybrid model’s performance is depending on the variety of data that are taken for forecasting.
Keywords: ARIMA, ANN, SVM, Time series, ARMA.
Scope of the Article: Time and Knowledge Management Tools