IBM Planning Analytics Workspace Release 2.0.58 news
IBM has released Planning Analytics AI forecasting. This has been a long-awaited feature, as Cognos Analytics has including several aspects of Artificial Intelligence for some time. The new feature can take a time-related set of data and choose from a number of forecasting methods to find the best-fit projection for the future with a confidence interval. This forecast can then be saved, for example as a version.
The forecasting algorithms discover and model trends, seasonality, and time dependence in data. The forecast is displayed as a continuation of historic data with confidence intervals highlighted.
(https://www.ibm.com/support/knowledgecenter/SSD29G_2.0.0/com.ibm.swg.ba.cognos.tm1_prism_gs.2.0.0.doc/c_paw_forcasting_nf.html (c) IBM)
Output from forecasting
The forecasting will look beyond the data displayed on the screen and needs at least twice the time period being forecasted and at least two seasons of data for seasonality.
The output from forecasting will contain:
- Prediction accuracy
- Preview chart
- The statistical details
Model estimation algorithms
- Smoothing equations
- Level smoothing equations
- Trend smoothing equations
- Seasonal smoothing equations
- One-step ahead
- k-step ahead
- Confidence bounds
The AI can simulate and choose the most optimal algorithms, or you may choose your own.
The introduction of AI into the already powerful TM1 product brings planning into the twenty-first century. Look forward to more AI embedded in Planning Analytics as IBM continues to integrate Watson-inspired AI into their products.