Statistical Method
New Time Series Update: Augmented Dickey-Fuller Test
User Benefit
Summary
Augmented Dickey-Fuller Test determines whether differencing makes the mean of the data stationary. This command is used to determine the nonseasonal differencing order when you analyze your time series data with an ARIMA model.
Statistical Method
New Time Series Update: Box-Cox Transformation
User Benefit
Box-Cox Transformation stabilizes the variance of a time series over time for better forecasting results.
Summary
Box-Cox Transformation is a technique available on the MSS Time Series menu. It is used to make the variance of transformed time series data unchanged over time.
Statistical Method
New Time Series Update: Forecasting Best ARIMA
User Benefit
The Forecasting Best ARIMA command will do the model selection analyses for the user automatically, saving them time from having to evaluate different ARIMA models themselves and selecting best model.
Summary
Forecasting Best ARIMA automatically forecasts the future values of a times series data set. The user can also select a simpler-but-equally-good alternative model to avoid any over-fitting problems. The user is empowered to decide whether the selected model can fit the data reasonably well and produce reliable forecasting results.
Visualizations
Graph Builder Update: New Graphs Added
- Line Plot
- Stacked Area Graph
- Pie Chart
User Benefit
These new graphs empower users to choose the best graph when utilizing Graph Builder for their analysis.
Summary
Graph Builder’s easy-to-browse gallery lets users seamlessly switch from one graph to the next using the same selection of data and without re-running your analysis. Users will now be able to select newly added graphs from Graph Builder: Line Plot, Stacked Area Graph, and a Pie Chart.
