Revelwood

Your SUPER-powered WP Engine Site

Machine Learning

Workday Adaptive Planning Tips & Tricks: Machine Learning Predictive Forecaster: Lever Sheets

by

The machine learning predictive forecaster in Workday Adaptive Planning allows users to leverage machine learning to make predictions for financial planning and analysis. After setting up the initial requirements for a predictive forecaster, including name, sheet, forecast version, start period, end period, accounts, reference data and algorithm users may choose to add a lever modeled sheet to enhance the prediction.

Step 1: Understanding Lever Sheets

First determine the forecast sheet to use. Next review the levels, accounts, and dimensions of the forecast sheet that will be incorporated into the lever sheet. Identify the smallest timer period represented by the forecast sheet. 

Lever sheets are designed to hold regressor data that influences the machine learning forecasts. These sheets can be used only with specific algorithms and the values that are inputted into a lever sheet directly impact the algorithm;s calculation, which is reflected in the forecast sheet. 

Step 2: Create Lever Sheet

Navigate to Modeling > Level Assigned Sheets > New Sheet > Modeled Sheet (blank modeled sheet) 

Under Columns and Levels add in a field called Account as a text field. You can also add in dimensions if those dimensions exist on the forecast sheet. Next, drag the timespan into the last row. Lastly, choose settings under sheet properties and select the time stratum with the smallest time period. 

Step 3: Entering Data

Lever sheets contain data that influence the machine learning forecasts for a specific sheet. The values entered in the lever sheet modify the algorithm’s calculations by levels, account, time period, and dimensions (if applicable). Choose the sheet used for levers and make sure the same version as the forecasted data is selected. Next choose add row and enter account, level, and dimension. Repeat to add all necessary rows. Enter data in the timer periods between start and end. Save the sheet. 

When creating the predictive forecaster make sure to select the lever sheet that aligns with your forecast. Only the algorithms AutoFit, Orbit DLT, and Prophet allow lever sheets to be added. This additional sheet can help users create an even more fine tuned prediction for data on standard or cube sheets.

Revelwood is an award-winning, Platinum Solution Provider for Workday Adaptive Planning. We build solutions for the Office of Finance that minimize your risk by seamlessly incorporating business analytics into your everyday thinking. By combining the software with our best practices and out-of-the-box applications, we help businesses achieve their full potential with Workday Adaptive Planning.

Read more Workday Adaptive Planning Tips & Tricks:

Workday Adaptive Planning User Interface Changes 2025 R1 

Workday Adaptive Planning Tips & Tricks: Write-back for OfficeConnect

Workday Adaptive Planning Tips & Tricks: New Feature – Long-Running Processes in Workday’s 2024R2 Release

Home » Machine Learning

Workday Adaptive Planning Tips & Tricks: Machine Learning Predictive Forecaster

by

Workday Adaptive Planning offers several features that allow users to achieve financial planning and analysis data solutions. One feature, predictive forecaster, leverages historical data to generate machine learning-based predictions within forecasted versions. With improved efficiency and accuracy, predictive forecasters can provide well-tuned predictions with the help of machine learning algorithms. In this blog post, we will delve into the steps needed to set up a predictive forecaster in Workday Adaptive Planning. 

Step 1: Permission Sets

In order to access the predictive forecaster, navigate to the “Permission Sets” tab of the “Administration” section. From there click edit (next to the Permission Sets of the accessing users). Finally, make sure the predictive forecaster is checked off. Now predictive forecaster will be available under “modeling.”

 

Step 2: Navigation 

In order to create a new predictive forecaster navigate to the “modeling” tab, choose “predictive forecast” and click “add new”.

Step 3: Necessary Requirements

To create the forecast, start by including a name and an optional description. Next, choose the sheet and the planned version that you want to populate with the data. Additionally, add the forecast start and end periods.

Step 4: Requirements continued

To continue filling out the necessary requirements select the accounts that you want to populate with the predicted data. There is the option to choose whether you want to populate based on levels or dimensions. Next for the “reference data” section, select the actual version that the algorithm will reference, as well as the start and end periods. Adaptive Planning requires users to maintain a 3:2 ratio of actuals to forecast.

Step 5: Algorithms

Finally, users must select which algorithm they want to deploy for their forecaster. AutoFit is the suggested algorithm because it selects the algorithm based on the seasonality of the data. Users also have the option to choose type of seasonality, add weekly seasonality, add multiplicative seasonality, to round values, and to add lever sheets. Several of these additional selections are only available for specific algorithms.

Predictive forecasting allows for users to leverage historical data to create well-defined predictions. The detailed criteria helps to create tailored data to the unique seasonality of the data. 

Revelwood is an award-winning, Platinum Solution Provider for Workday Adaptive Planning. We build solutions for the Office of Finance that minimize your risk by seamlessly incorporating business analytics into your everyday thinking. By combining the software with our best practices and out-of-the-box applications, we help businesses achieve their full potential with Workday Adaptive Planning.

Read more Workday Adaptive Planning Tips & Tricks:

Top 5 Workday Adaptive Planning Tips & Tricks from 2024

Workday Adaptive Planning Tips & Tricks: Archived Versions For Data Freeze 2024R2

Workday Adaptive Planning Tips & Tricks: Data Validation Rules in Modeled Sheets

Home » Machine Learning