Chapter 6 Conclusion

6.1 Linear Regression:

Simple and easy to interpret but limited in capturing complex patterns and sensitive to outliers.

6.2 Quadratic Regression:

Adds flexibility for slight non-linearity but may overfit and has issues with coefficient significance.

6.3 ARIMA:

Best captures trends and seasonality in time series data, offering the most accurate forecasts, albeit with greater complexity.

For a balanced trade-off between model simplicity and forecasting accuracy, the ARIMA(0,1,2) model emerges as a suitable choice. However, for slightly improved accuracy, the ARIMA(2,1,2) model can be considered, accepting the trade-off of increased complexity.