A Practical Approach to Timeseries Forecasting Using Python - Variations in SARIMA

The video tutorial explores the use of SARIMA models with different seasonal values, demonstrating how variations in seasonal order affect forecasting results. It highlights the impact of reducing seasonality, showing that when seasonality is minimized, the model behaves similarly to ARIMA. The tutorial applies these concepts to COVID data, emphasizing the importance of choosing the correct seasonal period. The project concludes with a summary of SARIMA variations and a preview of future projects.

What happens when a two-year seasonal difference is used in a SARIMA model?

• It has no effect on the model.

• It decreases the accuracy due to lack of data.

• It makes the model faster.

• It increases the accuracy of the model.

How does a six-month seasonal order affect the SARIMA model?

• It reduces the variation in results.

• It increases the variation in results.

• It makes the model behave like ARIMA.

• It eliminates seasonality completely.

What is the effect of using a two-month seasonal period in SARIMA?

• It has no effect on the model.

• It introduces more seasonality.

• It makes the model behave like ARIMA.

• It increases the model's complexity.

Why did the model show worse results with a two-year seasonal period?

• Because the model was not trained properly.

• Because the data had a two-year cycle.

• Because there was insufficient data for two years.

• Because the seasonal order was too short.

What is the key takeaway regarding the use of SARIMA models in this project?

• SARIMA models do not require seasonal data.

• Any seasonal period can be used interchangeably.

• The correct seasonal period is crucial for accurate forecasting.

• SARIMA models are always better than ARIMA.