Research Article
Geometric Brownian Motion-Based Time Series Modeling Methodology for Statistical Autocorrelated Process Control: Logarithmic Return Model
Table 1
Results of an exploratory study.
| S. No. | References | No. of datasets | Model accuracy (MAPE in %) | Running time (seconds of CPU time) | Total | Positive | GBM | LR | ARIMA | LR | ARIMA |
| 1 | Box et al. [24] | 20 | 16 | 13 | 0.42–9.17 | 0.42–8.21 | 0.08–0.26 | 4.80–5.96 | 2 | Brockwell and Davis [29] | 40 | 27 | 20 | 0.26–9.84 | 0.25–9.36 | 0.04–0.22 | 5.10–6.44 | 3 | Cryer and Chan [30] | 48 | 35 | 24 | 0.14–8.93 | 0.14–7.60 | 0.06–0.28 | 4.85–6.11 | 4 | Harvey [31] | 18 | 17 | 8 | 0.29–8.93 | 0.31–6.76 | 0.04–0.11 | 4.84–5.75 | 5 | Montgomery et al. [32] | 25 | 24 | 20 | 0.03–7.69 | 0.05–7.36 | 0.04–0.31 | 5.00–6.28 | | Total | 151 | 119 | 85 | | | | |
|
|
Note: number of datasets with positive data; number of positive datasets explored using GBM law. |