Author: Tanujit Chakraborty; Indrajit Ghosh
Title: Real-time forecasts and risk assessment of novel coronavirus (COVID-19) cases: A data-driven analysis Document date: 2020_4_14
ID: ba6mdgq3_22
Snippet: For the COVID-19 datasets, we propose a hybridization of stationary ARIMA and nonstationary WBF model to reduce the individual biases of the component models [24] . The COVID-19 cases datasets for five different countries are complex in nature. Thus, the ARIMA model fails to produce random errors or even nonstationary residual series, evident from Figure 1 . The behavior of the residual series generated by ARIMA is mostly oscillatory and periodic.....
Document: For the COVID-19 datasets, we propose a hybridization of stationary ARIMA and nonstationary WBF model to reduce the individual biases of the component models [24] . The COVID-19 cases datasets for five different countries are complex in nature. Thus, the ARIMA model fails to produce random errors or even nonstationary residual series, evident from Figure 1 . The behavior of the residual series generated by ARIMA is mostly oscillatory and periodic; thus, we choose the wavelet function to model the remaining series. Several hybrid models based on ARIMA and neural networks are available in the field of time series forecasting; see for example [35; 1; 12; 19; 8; 25] .
Search related documents:
Co phrase search for related documents- arima model and component model: 1, 2
- arima model and different country: 1
- arima model and hybrid model: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
- arima model and neural network: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19
- arima model and random error: 1
- arima model and residual series: 1, 2, 3
- arima model and time series forecasting: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22
- component model and hybrid model: 1, 2, 3, 4
- component model and neural network: 1, 2, 3, 4, 5, 6, 7, 8
- component model and residual series: 1
- different country and hybrid model: 1
- different country and neural network: 1
- hybrid model and neural network: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49
- hybrid model and random error: 1
- hybrid model and residual series: 1, 2
- hybrid model and time series forecasting: 1, 2, 3, 4, 5, 6
- individual bias and neural network: 1
- neural network and time series forecasting: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30
- neural network and wavelet function: 1
Co phrase search for related documents, hyperlinks ordered by date